US:CPS: EmploymentStatus classification:: cps2
bdanalytics
Date: (Fri) Jun 12, 2015
Introduction:
Data: Source: Training: https://courses.edx.org/asset-v1:MITx+15.071x_2a+2T2015+type@asset+block/CPSData.csv
New:
Time period:
Synopsis:
Based on analysis utilizing <> techniques,
[](.png)
Potential next steps include:
- Organization:
- Categorize by chunk
- Priority criteria:
- Ease of change
- Impacts report
- Cleans innards
- Bug report
- all chunks:
- at chunk-end rm(!glb_)
- manage.missing.data chunk:
- cleaner way to manage re-splitting of training vs. new entity
- extract.features chunk:
- Add n-grams for glb_txt_vars
- “RTextTools”, “tau”, “RWeka”, and “textcat” packages
- Convert user-specified mutate code to config specs
- Add n-grams for glb_txt_vars
- fit.models chunk:
- Prediction accuracy scatter graph:
- Add tiles (raw vs. PCA)
- Use shiny for drop-down of “important” features
Use plot.ly for interactive plots ?
- Change .fit suffix of model metrics to .mdl if it’s data independent (e.g. AIC, Adj.R.Squared - is it truly data independent ?, etc.)
- move model_type parameter to myfit_mdl before indep_vars_vctr (keep all model_* together)
- create a custom model for rpart that has minbucket as a tuning parameter
varImp for randomForest crashes in caret version:6.0.41 -> submit bug report
- Probability handling for multinomials vs. desired binomial outcome
- ROCR currently supports only evaluation of binary classification tasks (version 1.0.7)
extensions toward multiclass classification are scheduled for the next release
- Skip trControl.method=“cv” for dummy classifier ?
- Add custom model to caret for a dummy (baseline) classifier (binomial & multinomial) that generates proba/outcomes which mimics the freq distribution of glb_rsp_var values; Right now glb_dmy_glm_mdl always generates most frequent outcome in training data
glm_dmy_mdl should use the same method as glm_sel_mdl until custom dummy classifer is implemented
- fit.all.training chunk:
- myplot_prediction_classification: displays ‘x’ instead of ‘+’ when there are no prediction errors
- Compare glb_sel_mdl vs. glb_fin_mdl:
- varImp
- Prediction differences (shd be minimal ?)
- Move glb_analytics_diag_plots to mydsutils.R: (+) Easier to debug (-) Too many glb vars used
- Add print(ggplot.petrinet(glb_analytics_pn) + coord_flip()) at the end of every major chunk
- Parameterize glb_analytics_pn
- Move glb_impute_missing_data to mydsutils.R: (-) Too many glb vars used; glb_<>_df reassigned
- Replicate myfit_mdl_classification features in myfit_mdl_regression
- Do non-glm methods handle interaction terms ?
- f-score computation for classifiers should be summation across outcomes (not just the desired one ?)
- Add accuracy computation to glb_dmy_mdl in predict.data.new chunk
- Why does splitting fit.data.training.all chunk into separate chunks add an overhead of ~30 secs ? It’s not rbind b/c other chunks have lower elapsed time. Is it the number of plots ?
- Incorporate code chunks in print_sessionInfo
- Test against
- projects in github.com/bdanalytics
- lectures in jhu-datascience track
Analysis:
rm(list=ls())
set.seed(12345)
options(stringsAsFactors=FALSE)
source("~/Dropbox/datascience/R/myscript.R")
source("~/Dropbox/datascience/R/mydsutils.R")## Loading required package: caret
## Loading required package: lattice
## Loading required package: ggplot2source("~/Dropbox/datascience/R/myplot.R")
source("~/Dropbox/datascience/R/mypetrinet.R")
source("~/Dropbox/datascience/R/myplclust.R")
# Gather all package requirements here
suppressPackageStartupMessages(require(doMC))
registerDoMC(4) # max(length(glb_txt_vars), glb_n_cv_folds) + 1
#packageVersion("snow")
#require(sos); findFn("cosine", maxPages=2, sortby="MaxScore")
# Analysis control global variables
glb_trnng_url <- "https://courses.edx.org/asset-v1:MITx+15.071x_2a+2T2015+type@asset+block/CPSData.csv"
glb_newdt_url <- "<newdt_url>"
glb_out_pfx <- "cps2_"
glb_save_envir <- FALSE # or TRUE
glb_is_separate_newent_dataset <- FALSE # or TRUE
glb_split_entity_newent_datasets <- TRUE # or FALSE
glb_split_newdata_method <- "condition" # "condition" or "sample" or "copy"
glb_split_newdata_condition <- "is.na(EmploymentStatus)" # "<col_name> <condition_operator> <value>" # or NULL
glb_split_newdata_size_ratio <- 0.3 # > 0 & < 1
glb_split_sample.seed <- 123 # or any integer
glb_drop_vars <- c(NULL) # or c("<col_name>")
glb_max_fitent_obs <- NULL # or any integer
glb_is_regression <- FALSE; glb_is_classification <- TRUE; glb_is_binomial <- FALSE
glb_rsp_var_raw <- "EmploymentStatus"
# for classification, the response variable has to be a factor
glb_rsp_var <- "EmploymentStatus.fctr"
# if the response factor is based on numbers/logicals e.g (0/1 OR TRUE/FALSE vs. "A"/"B"),
# or contains spaces (e.g. "Not in Labor Force")
# caret predict(..., type="prob") crashes
glb_map_rsp_raw_to_var <- function(raw) {
#relevel(factor(ifelse(raw == 1, "Y", "N")), as.factor(c("Y", "N")), ref="N")
#as.factor(paste0("B", raw))
as.factor(gsub(" ", "\\.", raw))
}
glb_map_rsp_raw_to_var(
c("Retired", "Unemployed", "Disabled", "Not in Labor Force", "Employed"))## [1] Retired Unemployed Disabled
## [4] Not.in.Labor.Force Employed
## Levels: Disabled Employed Not.in.Labor.Force Retired Unemployedglb_map_rsp_var_to_raw <- function(var) {
#as.numeric(var) - 1
#as.numeric(var)
gsub("\\.", " ", levels(var)[as.numeric(var)])
#c(" <=50K", " >50K")[as.numeric(var)]
#c(FALSE, TRUE)[as.numeric(var)]
}
glb_map_rsp_var_to_raw(glb_map_rsp_raw_to_var(
c("Retired", "Unemployed", "Disabled", "Not in Labor Force", "Employed")))## [1] "Retired" "Unemployed" "Disabled"
## [4] "Not in Labor Force" "Employed"if ((glb_rsp_var != glb_rsp_var_raw) & is.null(glb_map_rsp_raw_to_var))
stop("glb_map_rsp_raw_to_var function expected")
glb_rsp_var_out <- paste0(glb_rsp_var, ".predict.") # model_id is appended later
# List info gathered for various columns
# <col_name>: <description>; <notes>
# PeopleInHousehold: The number of people in the interviewee's household.
# Region: The census region where the interviewee lives.
# State: The state where the interviewee lives.
# MetroAreaCode: A code that identifies the metropolitan area in which the interviewee lives (missing if the interviewee does not live in a metropolitan area). The mapping from codes to names of metropolitan areas is provided in the file MetroAreaCodes.csv.
# Age: The age, in years, of the interviewee. 80 represents people aged 80-84, and 85 represents people aged 85 and higher.
# Married: The marriage status of the interviewee.
# Sex: The sex of the interviewee.
# Education: The maximum level of education obtained by the interviewee.
# Race: The race of the interviewee.
# Hispanic: Whether the interviewee is of Hispanic ethnicity.
# CountryOfBirthCode: A code identifying the country of birth of the interviewee. The mapping from codes to names of countries is provided in the file CountryCodes.csv.
# Citizenship: The United States citizenship status of the interviewee.
# EmploymentStatus: The status of employment of the interviewee.
# Industry: The industry of employment of the interviewee (only available if they are employed).
# If multiple vars are parts of id, consider concatenating them to create one id var
# If glb_id_var == NULL, ".rownames <- row.names()" is the default
glb_id_var <- NULL # or c("<var1>")
glb_category_vars <- NULL # or c("<var1>", "<var2>")
glb_map_vars <- c("MetroAreaCode", "CountryOfBirthCode")
glb_map_urls <- list();
glb_map_urls[["MetroAreaCode"]] <- "https://courses.edx.org/asset-v1:MITx+15.071x_2a+2T2015+type@asset+block/MetroAreaCodes.csv"
glb_map_urls[["CountryOfBirthCode"]] <- "https://courses.edx.org/asset-v1:MITx+15.071x_2a+2T2015+type@asset+block/CountryCodes.csv"
glb_assign_pairs_lst <- NULL;
glb_assign_pairs_lst[["Married"]] <- list(from=c(NA),
to=c("NA.my"))
glb_assign_pairs_lst[["Education"]] <- list(from=c(NA),
to=c("NA.my"))
glb_assign_pairs_lst[["Industry"]] <- list(from=c(NA),
to=c("NA.my"))
glb_assign_pairs_lst[["MetroArea"]] <- list(from=c(NA),
to=c("NA.my"))
glb_assign_pairs_lst[["Country"]] <- list(from=c(NA),
to=c("NA.my"))
glb_assign_vars <- names(glb_assign_pairs_lst)
glb_date_vars <- NULL # or c("<date_var>")
glb_date_fmts <- list(); #glb_date_fmts[["<date_var>"]] <- "%m/%e/%y"
glb_date_tzs <- list(); #glb_date_tzs[["<date_var>"]] <- "America/New_York"
#grep("America/New", OlsonNames(), value=TRUE)
glb_txt_vars <- NULL # or c("<txt_var1>", "<txt_var2>")
#Sys.setlocale("LC_ALL", "C") # For english
glb_append_stop_words <- list()
# Remember to use unstemmed words
#orderBy(~ -cor.y.abs, subset(glb_feats_df, grepl("[HSA]\\.T\\.", id) & !is.na(cor.high.X)))
#dsp_obs(Headline.contains="polit")
#subset(glb_allobs_df, H.T.compani > 0)[, c("UniqueID", "Headline", "H.T.compani")]
# glb_append_stop_words[["<txt_var1>"]] <- c(NULL
# # ,"<word1>" # <reason1>
# )
#subset(glb_allobs_df, S.T.newyorktim > 0)[, c("UniqueID", "Snippet", "S.T.newyorktim")]
#glb_txt_lst[["Snippet"]][which(glb_allobs_df$UniqueID %in% c(8394, 8317, 8339, 8350, 8307))]
glb_important_terms <- list()
# Remember to use stemmed terms
glb_sprs_thresholds <- NULL # or c(0.988, 0.970, 0.970) # Generates 29, 22, 22 terms
# Properties:
# numrows(glb_feats_df) << numrows(glb_fitobs_df)
# Select terms that appear in at least 0.2 * O(FP/FN(glb_OOBobs_df))
# numrows(glb_OOBobs_df) = 1.1 * numrows(glb_newobs_df)
names(glb_sprs_thresholds) <- glb_txt_vars
glb_log_vars <- NULL # or c("<numeric_var1>", "<numeric_var2>")
# List transformed vars
glb_exclude_vars_as_features <- c("State.fctr", "MetroArea.my.fctr", "Country.my.fctr")
if (glb_rsp_var_raw != glb_rsp_var)
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
glb_rsp_var_raw)
# List feats that shd be excluded due to known causation by prediction variable
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
c(NULL)) # or c("<col_name>")
glb_impute_na_data <- FALSE # or TRUE
glb_mice_complete.seed <- 144 # or any integer
glb_cluster <- FALSE # or TRUE
glb_interaction_only_features <- NULL # or ???
glb_models_lst <- list(); glb_models_df <- data.frame()
# Regression
if (glb_is_regression)
glb_models_method_vctr <- c("lm", "glm", "bayesglm", "rpart", "rf") else
# Classification
if (glb_is_binomial)
glb_models_method_vctr <- c("glm", "bayesglm", "rpart", "rf") else
#glb_models_method_vctr <- c("rpart", "rf")
glb_models_method_vctr <- c("rpart")
# Baseline prediction model feature(s)
glb_Baseline_mdl_var <- NULL # or c("<col_name>")
glb_model_metric_terms <- NULL # or matrix(c(
# 0,1,2,3,4,
# 2,0,1,2,3,
# 4,2,0,1,2,
# 6,4,2,0,1,
# 8,6,4,2,0
# ), byrow=TRUE, nrow=5)
glb_model_metric <- NULL # or "<metric_name>"
glb_model_metric_maximize <- NULL # or FALSE (TRUE is not the default for both classification & regression)
glb_model_metric_smmry <- NULL # or function(data, lev=NULL, model=NULL) {
# confusion_mtrx <- t(as.matrix(confusionMatrix(data$pred, data$obs)))
# #print(confusion_mtrx)
# #print(confusion_mtrx * glb_model_metric_terms)
# metric <- sum(confusion_mtrx * glb_model_metric_terms) / nrow(data)
# names(metric) <- glb_model_metric
# return(metric)
# }
glb_tune_models_df <-
rbind(
#data.frame(parameter="cp", min=0.00005, max=0.00005, by=0.000005),
#seq(from=0.01, to=0.01, by=0.01)
#data.frame(parameter="mtry", min=080, max=100, by=10),
#data.frame(parameter="mtry", min=08, max=10, by=1),
data.frame(parameter="dummy", min=2, max=4, by=1)
)
# or NULL
glb_n_cv_folds <- 3 # or NULL
glb_clf_proba_threshold <- NULL # 0.5
# Model selection criteria
if (glb_is_regression)
glb_model_evl_criteria <- c("min.RMSE.OOB", "max.R.sq.OOB", "max.Adj.R.sq.fit")
if (glb_is_classification) {
if (glb_is_binomial)
glb_model_evl_criteria <-
c("max.Accuracy.OOB", "max.auc.OOB", "max.Kappa.OOB", "min.aic.fit") else
glb_model_evl_criteria <- c("max.Accuracy.OOB", "max.Kappa.OOB")
}
glb_sel_mdl_id <- NULL # or "<model_id_prefix>.<model_method>"
glb_fin_mdl_id <- glb_sel_mdl_id # or "Final"
# Depict process
glb_analytics_pn <- petrinet(name="glb_analytics_pn",
trans_df=data.frame(id=1:6,
name=c("data.training.all","data.new",
"model.selected","model.final",
"data.training.all.prediction","data.new.prediction"),
x=c( -5,-5,-15,-25,-25,-35),
y=c( -5, 5, 0, 0, -5, 5)
),
places_df=data.frame(id=1:4,
name=c("bgn","fit.data.training.all","predict.data.new","end"),
x=c( -0, -20, -30, -40),
y=c( 0, 0, 0, 0),
M0=c( 3, 0, 0, 0)
),
arcs_df=data.frame(
begin=c("bgn","bgn","bgn",
"data.training.all","model.selected","fit.data.training.all",
"fit.data.training.all","model.final",
"data.new","predict.data.new",
"data.training.all.prediction","data.new.prediction"),
end =c("data.training.all","data.new","model.selected",
"fit.data.training.all","fit.data.training.all","model.final",
"data.training.all.prediction","predict.data.new",
"predict.data.new","data.new.prediction",
"end","end")
))
#print(ggplot.petrinet(glb_analytics_pn))
print(ggplot.petrinet(glb_analytics_pn) + coord_flip())## Loading required package: grid
glb_analytics_avl_objs <- NULL
glb_chunks_df <- myadd_chunk(NULL, "import.data")## label step_major step_minor bgn end elapsed
## 1 import.data 1 0 12.518 NA NAStep 1.0: import data
chunk option: eval=
glb_trnobs_df <- myimport_data(url=glb_trnng_url, comment="glb_trnobs_df",
force_header=TRUE)## [1] "Reading file ./data/CPSData.csv..."
## [1] "dimensions of data in ./data/CPSData.csv: 131,302 rows x 14 cols"
## PeopleInHousehold Region State MetroAreaCode Age Married Sex
## 1 1 South Alabama 26620 85 Widowed Female
## 2 3 South Alabama 13820 21 Never Married Male
## 3 3 South Alabama 13820 37 Never Married Female
## 4 3 South Alabama 13820 18 Never Married Male
## 5 3 South Alabama 26620 52 Widowed Female
## 6 3 South Alabama 26620 24 Never Married Male
## Education Race Hispanic CountryOfBirthCode Citizenship
## 1 Associate degree White 0 57 Citizen, Native
## 2 High school Black 0 57 Citizen, Native
## 3 High school Black 0 57 Citizen, Native
## 4 No high school diploma Black 0 57 Citizen, Native
## 5 Associate degree White 0 57 Citizen, Native
## 6 Bachelor's degree White 0 57 Citizen, Native
## EmploymentStatus Industry
## 1 Retired <NA>
## 2 Unemployed Professional and business services
## 3 Disabled <NA>
## 4 Not in Labor Force <NA>
## 5 Employed Professional and business services
## 6 Employed Educational and health services
## PeopleInHousehold Region State MetroAreaCode Age
## 4535 7 South Arkansas 30780 6
## 20007 1 West Colorado 14500 57
## 66863 3 South Mississippi NA 48
## 95549 6 South Oklahoma NA 17
## 96594 1 West Oregon 38900 85
## 129953 2 West Wyoming NA 54
## Married Sex Education Race Hispanic
## 4535 <NA> Female <NA> Black 0
## 20007 Divorced Male Master's degree White 0
## 66863 Never Married Female No high school diploma Black 0
## 95549 Never Married Male No high school diploma White 0
## 96594 Widowed Female No high school diploma White 0
## 129953 Married Male High school White 0
## CountryOfBirthCode Citizenship EmploymentStatus Industry
## 4535 57 Citizen, Native <NA> <NA>
## 20007 57 Citizen, Native Employed Financial
## 66863 57 Citizen, Native Disabled <NA>
## 95549 57 Citizen, Native Not in Labor Force <NA>
## 96594 57 Citizen, Native Retired <NA>
## 129953 57 Citizen, Native Employed Manufacturing
## PeopleInHousehold Region State MetroAreaCode Age Married
## 131297 5 West Wyoming NA 14 <NA>
## 131298 5 West Wyoming NA 17 Never Married
## 131299 5 West Wyoming NA 37 Divorced
## 131300 3 West Wyoming NA 58 Married
## 131301 3 West Wyoming NA 53 Married
## 131302 3 West Wyoming NA 14 <NA>
## Sex Education Race Hispanic CountryOfBirthCode
## 131297 Male <NA> White 0 57
## 131298 Male No high school diploma White 0 57
## 131299 Male High school White 0 57
## 131300 Male Bachelor's degree White 0 57
## 131301 Female Associate degree White 0 57
## 131302 Female <NA> White 0 57
## Citizenship EmploymentStatus Industry
## 131297 Citizen, Native <NA> <NA>
## 131298 Citizen, Native Not in Labor Force <NA>
## 131299 Citizen, Native Employed Mining
## 131300 Citizen, Native Employed Financial
## 131301 Citizen, Native Not in Labor Force <NA>
## 131302 Citizen, Native <NA> <NA>
## 'data.frame': 131302 obs. of 14 variables:
## $ PeopleInHousehold : int 1 3 3 3 3 3 3 2 2 2 ...
## $ Region : chr "South" "South" "South" "South" ...
## $ State : chr "Alabama" "Alabama" "Alabama" "Alabama" ...
## $ MetroAreaCode : int 26620 13820 13820 13820 26620 26620 26620 33660 33660 26620 ...
## $ Age : int 85 21 37 18 52 24 26 71 43 52 ...
## $ Married : chr "Widowed" "Never Married" "Never Married" "Never Married" ...
## $ Sex : chr "Female" "Male" "Female" "Male" ...
## $ Education : chr "Associate degree" "High school" "High school" "No high school diploma" ...
## $ Race : chr "White" "Black" "Black" "Black" ...
## $ Hispanic : int 0 0 0 0 0 0 0 0 0 0 ...
## $ CountryOfBirthCode: int 57 57 57 57 57 57 57 57 57 57 ...
## $ Citizenship : chr "Citizen, Native" "Citizen, Native" "Citizen, Native" "Citizen, Native" ...
## $ EmploymentStatus : chr "Retired" "Unemployed" "Disabled" "Not in Labor Force" ...
## $ Industry : chr NA "Professional and business services" NA NA ...
## - attr(*, "comment")= chr "glb_trnobs_df"
## NULL# glb_trnobs_df <- data.frame()
# for (symbol in c("Boeing", "CocaCola", "GE", "IBM", "ProcterGamble")) {
# sym_trnobs_df <-
# myimport_data(url=gsub("IBM", symbol, glb_trnng_url), comment="glb_trnobs_df",
# force_header=TRUE)
# sym_trnobs_df$Symbol <- symbol
# glb_trnobs_df <- myrbind_df(glb_trnobs_df, sym_trnobs_df)
# }
if (glb_is_separate_newent_dataset) {
glb_newobs_df <- myimport_data(url=glb_newdt_url, comment="glb_newobs_df",
force_header=TRUE)
# To make plots / stats / checks easier in chunk:inspectORexplore.data
glb_allobs_df <- myrbind_df(glb_trnobs_df, glb_newobs_df);
comment(glb_allobs_df) <- "glb_allobs_df"
} else {
glb_allobs_df <- glb_trnobs_df; comment(glb_allobs_df) <- "glb_allobs_df"
if (!glb_split_entity_newent_datasets) {
stop("Not implemented yet")
glb_newobs_df <- glb_trnobs_df[sample(1:nrow(glb_trnobs_df),
max(2, nrow(glb_trnobs_df) / 1000)),]
} else if (glb_split_newdata_method == "condition") {
glb_newobs_df <- do.call("subset",
list(glb_trnobs_df, parse(text=glb_split_newdata_condition)))
glb_trnobs_df <- do.call("subset",
list(glb_trnobs_df, parse(text=paste0("!(",
glb_split_newdata_condition,
")"))))
} else if (glb_split_newdata_method == "sample") {
require(caTools)
set.seed(glb_split_sample.seed)
split <- sample.split(glb_trnobs_df[, glb_rsp_var_raw],
SplitRatio=(1-glb_split_newdata_size_ratio))
glb_newobs_df <- glb_trnobs_df[!split, ]
glb_trnobs_df <- glb_trnobs_df[split ,]
} else if (glb_split_newdata_method == "copy") {
glb_trnobs_df <- glb_allobs_df
comment(glb_trnobs_df) <- "glb_trnobs_df"
glb_newobs_df <- glb_allobs_df
comment(glb_newobs_df) <- "glb_newobs_df"
} else stop("glb_split_newdata_method should be %in% c('condition', 'sample', 'copy')")
comment(glb_newobs_df) <- "glb_newobs_df"
myprint_df(glb_newobs_df)
str(glb_newobs_df)
if (glb_split_entity_newent_datasets) {
myprint_df(glb_trnobs_df)
str(glb_trnobs_df)
}
} ## PeopleInHousehold Region State MetroAreaCode Age Married Sex
## 14 4 South Alabama 26620 2 <NA> Female
## 15 4 South Alabama 26620 4 <NA> Male
## 18 2 South Alabama 13820 13 <NA> Female
## 28 3 South Alabama 33860 2 <NA> Female
## 35 6 South Alabama 33860 3 <NA> Female
## 36 6 South Alabama 33860 11 <NA> Female
## Education Race Hispanic CountryOfBirthCode Citizenship
## 14 <NA> White 0 57 Citizen, Native
## 15 <NA> White 0 57 Citizen, Native
## 18 <NA> Black 0 57 Citizen, Native
## 28 <NA> White 0 57 Citizen, Native
## 35 <NA> Black 0 57 Citizen, Native
## 36 <NA> Black 0 57 Citizen, Native
## EmploymentStatus Industry
## 14 <NA> <NA>
## 15 <NA> <NA>
## 18 <NA> <NA>
## 28 <NA> <NA>
## 35 <NA> <NA>
## 36 <NA> <NA>
## PeopleInHousehold Region State MetroAreaCode Age Married
## 229 3 South Alabama 13820 6 <NA>
## 22928 3 Northeast Connecticut 76450 13 <NA>
## 51143 5 South Louisiana 35380 2 <NA>
## 51529 4 South Louisiana 43340 11 <NA>
## 52960 5 Northeast Maine NA 5 <NA>
## 61204 3 Midwest Michigan 19820 0 <NA>
## Sex Education Race Hispanic CountryOfBirthCode Citizenship
## 229 Female <NA> Black 0 57 Citizen, Native
## 22928 Female <NA> White 1 57 Citizen, Native
## 51143 Female <NA> White 0 57 Citizen, Native
## 51529 Female <NA> Black 0 57 Citizen, Native
## 52960 Female <NA> White 0 57 Citizen, Native
## 61204 Male <NA> White 0 57 Citizen, Native
## EmploymentStatus Industry
## 229 <NA> <NA>
## 22928 <NA> <NA>
## 51143 <NA> <NA>
## 51529 <NA> <NA>
## 52960 <NA> <NA>
## 61204 <NA> <NA>
## PeopleInHousehold Region State MetroAreaCode Age Married Sex
## 131282 5 West Wyoming NA 4 <NA> Female
## 131283 5 West Wyoming NA 9 <NA> Female
## 131285 2 West Wyoming NA 21 Married Male
## 131296 5 West Wyoming NA 10 <NA> Female
## 131297 5 West Wyoming NA 14 <NA> Male
## 131302 3 West Wyoming NA 14 <NA> Female
## Education Race Hispanic CountryOfBirthCode Citizenship
## 131282 <NA> White 0 57 Citizen, Native
## 131283 <NA> White 0 57 Citizen, Native
## 131285 High school White 1 57 Citizen, Native
## 131296 <NA> White 0 57 Citizen, Native
## 131297 <NA> White 0 57 Citizen, Native
## 131302 <NA> White 0 57 Citizen, Native
## EmploymentStatus Industry
## 131282 <NA> <NA>
## 131283 <NA> <NA>
## 131285 <NA> <NA>
## 131296 <NA> <NA>
## 131297 <NA> <NA>
## 131302 <NA> <NA>
## 'data.frame': 25789 obs. of 14 variables:
## $ PeopleInHousehold : int 4 4 2 3 6 6 2 4 3 3 ...
## $ Region : chr "South" "South" "South" "South" ...
## $ State : chr "Alabama" "Alabama" "Alabama" "Alabama" ...
## $ MetroAreaCode : int 26620 26620 13820 33860 33860 33860 26620 33660 13820 13820 ...
## $ Age : int 2 4 13 2 3 11 5 14 5 11 ...
## $ Married : chr NA NA NA NA ...
## $ Sex : chr "Female" "Male" "Female" "Female" ...
## $ Education : chr NA NA NA NA ...
## $ Race : chr "White" "White" "Black" "White" ...
## $ Hispanic : int 0 0 0 0 0 0 0 0 0 0 ...
## $ CountryOfBirthCode: int 57 57 57 57 57 57 57 57 57 57 ...
## $ Citizenship : chr "Citizen, Native" "Citizen, Native" "Citizen, Native" "Citizen, Native" ...
## $ EmploymentStatus : chr NA NA NA NA ...
## $ Industry : chr NA NA NA NA ...
## - attr(*, "comment")= chr "glb_newobs_df"
## PeopleInHousehold Region State MetroAreaCode Age Married Sex
## 1 1 South Alabama 26620 85 Widowed Female
## 2 3 South Alabama 13820 21 Never Married Male
## 3 3 South Alabama 13820 37 Never Married Female
## 4 3 South Alabama 13820 18 Never Married Male
## 5 3 South Alabama 26620 52 Widowed Female
## 6 3 South Alabama 26620 24 Never Married Male
## Education Race Hispanic CountryOfBirthCode Citizenship
## 1 Associate degree White 0 57 Citizen, Native
## 2 High school Black 0 57 Citizen, Native
## 3 High school Black 0 57 Citizen, Native
## 4 No high school diploma Black 0 57 Citizen, Native
## 5 Associate degree White 0 57 Citizen, Native
## 6 Bachelor's degree White 0 57 Citizen, Native
## EmploymentStatus Industry
## 1 Retired <NA>
## 2 Unemployed Professional and business services
## 3 Disabled <NA>
## 4 Not in Labor Force <NA>
## 5 Employed Professional and business services
## 6 Employed Educational and health services
## PeopleInHousehold Region State MetroAreaCode Age
## 42850 2 Midwest Indiana 26900 73
## 59500 1 Northeast Massachusetts 71650 85
## 84212 3 Northeast New York 35620 51
## 92491 2 Midwest Ohio 18140 25
## 124956 4 West Washington NA 18
## 126712 4 South West Virginia 16620 29
## Married Sex Education Race Hispanic
## 42850 Married Female High school White 0
## 59500 Widowed Female High school White 0
## 84212 Married Male Bachelor's degree White 1
## 92491 Married Female Bachelor's degree White 0
## 124956 Never Married Female High school Black 1
## 126712 Married Male High school White 0
## CountryOfBirthCode Citizenship EmploymentStatus
## 42850 57 Citizen, Native Retired
## 59500 57 Citizen, Native Retired
## 84212 57 Citizen, Native Employed
## 92491 57 Citizen, Native Not in Labor Force
## 124956 57 Citizen, Native Not in Labor Force
## 126712 57 Citizen, Native Employed
## Industry
## 42850 <NA>
## 59500 <NA>
## 84212 Public administration
## 92491 <NA>
## 124956 <NA>
## 126712 Mining
## PeopleInHousehold Region State MetroAreaCode Age Married
## 131294 2 West Wyoming NA 27 Never Married
## 131295 5 West Wyoming NA 39 Divorced
## 131298 5 West Wyoming NA 17 Never Married
## 131299 5 West Wyoming NA 37 Divorced
## 131300 3 West Wyoming NA 58 Married
## 131301 3 West Wyoming NA 53 Married
## Sex Education Race Hispanic CountryOfBirthCode
## 131294 Male High school White 0 57
## 131295 Female Associate degree White 0 57
## 131298 Male No high school diploma White 0 57
## 131299 Male High school White 0 57
## 131300 Male Bachelor's degree White 0 57
## 131301 Female Associate degree White 0 57
## Citizenship EmploymentStatus
## 131294 Citizen, Native Unemployed
## 131295 Citizen, Native Not in Labor Force
## 131298 Citizen, Native Not in Labor Force
## 131299 Citizen, Native Employed
## 131300 Citizen, Native Employed
## 131301 Citizen, Native Not in Labor Force
## Industry
## 131294 Professional and business services
## 131295 <NA>
## 131298 <NA>
## 131299 Mining
## 131300 Financial
## 131301 <NA>
## 'data.frame': 105513 obs. of 14 variables:
## $ PeopleInHousehold : int 1 3 3 3 3 3 3 2 2 2 ...
## $ Region : chr "South" "South" "South" "South" ...
## $ State : chr "Alabama" "Alabama" "Alabama" "Alabama" ...
## $ MetroAreaCode : int 26620 13820 13820 13820 26620 26620 26620 33660 33660 26620 ...
## $ Age : int 85 21 37 18 52 24 26 71 43 52 ...
## $ Married : chr "Widowed" "Never Married" "Never Married" "Never Married" ...
## $ Sex : chr "Female" "Male" "Female" "Male" ...
## $ Education : chr "Associate degree" "High school" "High school" "No high school diploma" ...
## $ Race : chr "White" "Black" "Black" "Black" ...
## $ Hispanic : int 0 0 0 0 0 0 0 0 0 0 ...
## $ CountryOfBirthCode: int 57 57 57 57 57 57 57 57 57 57 ...
## $ Citizenship : chr "Citizen, Native" "Citizen, Native" "Citizen, Native" "Citizen, Native" ...
## $ EmploymentStatus : chr "Retired" "Unemployed" "Disabled" "Not in Labor Force" ...
## $ Industry : chr NA "Professional and business services" NA NA ...if ((num_nas <- sum(is.na(glb_trnobs_df[, glb_rsp_var_raw]))) > 0)
stop("glb_trnobs_df$", glb_rsp_var_raw, " contains NAs for ", num_nas, " obs")
if (nrow(glb_trnobs_df) == nrow(glb_allobs_df))
warning("glb_trnobs_df same as glb_allobs_df")
if (nrow(glb_newobs_df) == nrow(glb_allobs_df))
warning("glb_newobs_df same as glb_allobs_df")
if (length(glb_drop_vars) > 0) {
warning("dropping vars: ", paste0(glb_drop_vars, collapse=", "))
glb_allobs_df <- glb_allobs_df[, setdiff(names(glb_allobs_df), glb_drop_vars)]
glb_trnobs_df <- glb_trnobs_df[, setdiff(names(glb_trnobs_df), glb_drop_vars)]
glb_newobs_df <- glb_newobs_df[, setdiff(names(glb_newobs_df), glb_drop_vars)]
}
#stop(here"); sav_allobs_df <- glb_allobs_df # glb_allobs_df <- sav_allobs_df
# Check for duplicates in glb_id_var
if (length(glb_id_var) == 0) {
warning("using .rownames as identifiers for observations")
glb_allobs_df$.rownames <- rownames(glb_allobs_df)
glb_trnobs_df$.rownames <- rownames(glb_trnobs_df)
glb_newobs_df$.rownames <- rownames(glb_newobs_df)
glb_id_var <- ".rownames"
}## Warning: using .rownames as identifiers for observationsif (sum(duplicated(glb_allobs_df[, glb_id_var, FALSE])) > 0)
stop(glb_id_var, " duplicated in glb_allobs_df")
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, glb_id_var)
# Combine trnent & newent into glb_allobs_df for easier manipulation
glb_trnobs_df$.src <- "Train"; glb_newobs_df$.src <- "Test";
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, ".src")
glb_allobs_df <- myrbind_df(glb_trnobs_df, glb_newobs_df)
comment(glb_allobs_df) <- "glb_allobs_df"
glb_allobs_df <- orderBy(reformulate(glb_id_var), glb_allobs_df)
glb_trnobs_df <- glb_newobs_df <- NULL
glb_chunks_df <- myadd_chunk(glb_chunks_df, "inspect.data", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 1 import.data 1 0 12.518 19.517 6.999
## 2 inspect.data 2 0 19.517 NA NAStep 2.0: inspect data
#print(str(glb_allobs_df))
#View(glb_allobs_df)
dsp_class_dstrb <- function(var) {
xtab_df <- mycreate_xtab_df(glb_allobs_df, c(".src", var))
rownames(xtab_df) <- xtab_df$.src
xtab_df <- subset(xtab_df, select=-.src)
print(xtab_df)
print(xtab_df / rowSums(xtab_df, na.rm=TRUE))
}
# Performed repeatedly in other chunks
glb_chk_data <- function() {
# Histogram of predictor in glb_trnobs_df & glb_newobs_df
print(myplot_histogram(glb_allobs_df, glb_rsp_var_raw) + facet_wrap(~ .src))
if (glb_is_classification)
dsp_class_dstrb(var=ifelse(glb_rsp_var %in% names(glb_allobs_df),
glb_rsp_var, glb_rsp_var_raw))
mycheck_problem_data(glb_allobs_df)
}
glb_chk_data()## Warning in myplot_histogram(glb_allobs_df, glb_rsp_var_raw): converting
## EmploymentStatus to class:factor## Loading required package: reshape2
## EmploymentStatus.Disabled EmploymentStatus.Employed
## Test NA NA
## Train 5712 61733
## EmploymentStatus.Not in Labor Force EmploymentStatus.Retired
## Test NA NA
## Train 15246 18619
## EmploymentStatus.Unemployed EmploymentStatus.NA
## Test NA 25789
## Train 4203 NA
## EmploymentStatus.Disabled EmploymentStatus.Employed
## Test NA NA
## Train 0.05413551 0.5850748
## EmploymentStatus.Not in Labor Force EmploymentStatus.Retired
## Test NA NA
## Train 0.144494 0.1764617
## EmploymentStatus.Unemployed EmploymentStatus.NA
## Test NA 1
## Train 0.03983395 NA
## [1] "numeric data missing in glb_allobs_df: "
## MetroAreaCode
## 34238
## [1] "numeric data w/ 0s in glb_allobs_df: "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in glb_allobs_df: "
## named integer(0)
## [1] "numeric data w/ NaNs in glb_allobs_df: "
## named integer(0)
## [1] "string data missing in glb_allobs_df: "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames
## NA 0# Create new features that help diagnostics
if (!is.null(glb_map_rsp_raw_to_var)) {
glb_allobs_df[, glb_rsp_var] <-
glb_map_rsp_raw_to_var(glb_allobs_df[, glb_rsp_var_raw])
mycheck_map_results(mapd_df=glb_allobs_df,
from_col_name=glb_rsp_var_raw, to_col_name=glb_rsp_var)
if (glb_is_classification) dsp_class_dstrb(glb_rsp_var)
}## Loading required package: sqldf
## Loading required package: gsubfn
## Loading required package: proto
## Loading required package: RSQLite
## Loading required package: DBI
## Loading required package: tcltk## EmploymentStatus EmploymentStatus.fctr .n
## 1 Employed Employed 61733
## 2 <NA> <NA> 25789
## 3 Retired Retired 18619
## 4 Not in Labor Force Not.in.Labor.Force 15246
## 5 Disabled Disabled 5712
## 6 Unemployed Unemployed 4203## Warning in loop_apply(n, do.ply): Removed 1 rows containing missing values
## (position_stack).
## EmploymentStatus.fctr.Disabled EmploymentStatus.fctr.Employed
## Test NA NA
## Train 5712 61733
## EmploymentStatus.fctr.Not.in.Labor.Force
## Test NA
## Train 15246
## EmploymentStatus.fctr.Retired EmploymentStatus.fctr.Unemployed
## Test NA NA
## Train 18619 4203
## EmploymentStatus.fctr.NA
## Test 25789
## Train NA
## EmploymentStatus.fctr.Disabled EmploymentStatus.fctr.Employed
## Test NA NA
## Train 0.05413551 0.5850748
## EmploymentStatus.fctr.Not.in.Labor.Force
## Test NA
## Train 0.144494
## EmploymentStatus.fctr.Retired EmploymentStatus.fctr.Unemployed
## Test NA NA
## Train 0.1764617 0.03983395
## EmploymentStatus.fctr.NA
## Test 1
## Train NA# Convert dates to numbers
# typically, dates come in as chars;
# so this must be done before converting chars to factors
myextract_dates_df <- function(df, vars, id_vars, rsp_var) {
keep_feats <- c(NULL)
for (var in vars) {
dates_df <- df[, id_vars, FALSE]
dates_df[, rsp_var] <- df[, rsp_var, FALSE]
#dates_df <- data.frame(.date=strptime(df[, var], "%Y-%m-%d %H:%M:%S"))
dates_df <- cbind(dates_df, data.frame(.date=strptime(df[, var],
glb_date_fmts[[var]], tz=glb_date_tzs[[var]])))
# print(dates_df[is.na(dates_df$.date), c("ID", "Arrest.fctr", ".date")])
# print(glb_allobs_df[is.na(dates_df$.date), c("ID", "Arrest.fctr", "Date")])
# print(head(glb_allobs_df[grepl("4/7/02 .:..", glb_allobs_df$Date), c("ID", "Arrest.fctr", "Date")]))
# print(head(strptime(glb_allobs_df[grepl("4/7/02 .:..", glb_allobs_df$Date), "Date"], "%m/%e/%y %H:%M"))
# Wrong data during EST->EDT transition
# tmp <- strptime("4/7/02 2:00","%m/%e/%y %H:%M:%S"); print(tmp); print(is.na(tmp))
# dates_df[dates_df$ID == 2068197, .date] <- tmp
# grep("(.*?) 2:(.*)", glb_allobs_df[is.na(dates_df$.date), "Date"], value=TRUE)
# dates_df[is.na(dates_df$.date), ".date"] <-
# data.frame(.date=strptime(gsub("(.*?) 2:(.*)", "\\1 3:\\2",
# glb_allobs_df[is.na(dates_df$.date), "Date"]), "%m/%e/%y %H:%M"))$.date
if (sum(is.na(dates_df$.date)) > 0) {
stop("NA POSIX dates for ", var)
print(df[is.na(dates_df$.date), c(id_vars, rsp_var, var)])
}
.date <- dates_df$.date
dates_df[, paste0(var, ".POSIX")] <- .date
dates_df[, paste0(var, ".year")] <- as.numeric(format(.date, "%Y"))
dates_df[, paste0(var, ".year.fctr")] <- as.factor(format(.date, "%Y"))
dates_df[, paste0(var, ".month")] <- as.numeric(format(.date, "%m"))
dates_df[, paste0(var, ".month.fctr")] <- as.factor(format(.date, "%m"))
dates_df[, paste0(var, ".date")] <- as.numeric(format(.date, "%d"))
dates_df[, paste0(var, ".date.fctr")] <-
cut(as.numeric(format(.date, "%d")), 5) # by month week
dates_df[, paste0(var, ".juliandate")] <- as.numeric(format(.date, "%j"))
# wkday Sun=0; Mon=1; ...; Sat=6
dates_df[, paste0(var, ".wkday")] <- as.numeric(format(.date, "%w"))
dates_df[, paste0(var, ".wkday.fctr")] <- as.factor(format(.date, "%w"))
# Get US Federal Holidays for relevant years
require(XML)
doc.html = htmlTreeParse('http://about.usps.com/news/events-calendar/2012-federal-holidays.htm', useInternal = TRUE)
# # Extract all the paragraphs (HTML tag is p, starting at
# # the root of the document). Unlist flattens the list to
# # create a character vector.
# doc.text = unlist(xpathApply(doc.html, '//p', xmlValue))
# # Replace all \n by spaces
# doc.text = gsub('\\n', ' ', doc.text)
# # Join all the elements of the character vector into a single
# # character string, separated by spaces
# doc.text = paste(doc.text, collapse = ' ')
# parse the tree by tables
txt <- unlist(strsplit(xpathSApply(doc.html, "//*/table", xmlValue), "\n"))
# do some clean up with regular expressions
txt <- grep("day, ", txt, value=TRUE)
txt <- trimws(gsub("(.*?)day, (.*)", "\\2", txt))
# txt <- gsub("\t","",txt)
# txt <- sub("^[[:space:]]*(.*?)[[:space:]]*$", "\\1", txt, perl=TRUE)
# txt <- txt[!(txt %in% c("", "|"))]
hldays <- strptime(paste(txt, ", 2012", sep=""), "%B %e, %Y")
dates_df[, paste0(var, ".hlday")] <-
ifelse(format(.date, "%Y-%m-%d") %in% hldays, 1, 0)
# NYState holidays 1.9., 13.10., 11.11., 27.11., 25.12.
dates_df[, paste0(var, ".wkend")] <- as.numeric(
(dates_df[, paste0(var, ".wkday")] %in% c(0, 6)) |
dates_df[, paste0(var, ".hlday")] )
dates_df[, paste0(var, ".hour")] <- as.numeric(format(.date, "%H"))
dates_df[, paste0(var, ".hour.fctr")] <-
if (length(unique(vals <- as.numeric(format(.date, "%H")))) <= 1)
vals else cut(vals, 3) # by work-shift
dates_df[, paste0(var, ".minute")] <- as.numeric(format(.date, "%M"))
dates_df[, paste0(var, ".minute.fctr")] <-
if (length(unique(vals <- as.numeric(format(.date, "%M")))) <= 1)
vals else cut(vals, 4) # by quarter-hours
dates_df[, paste0(var, ".second")] <- as.numeric(format(.date, "%S"))
dates_df[, paste0(var, ".second.fctr")] <-
if (length(unique(vals <- as.numeric(format(.date, "%S")))) <= 1)
vals else cut(vals, 4) # by quarter-minutes
dates_df[, paste0(var, ".day.minutes")] <-
60 * dates_df[, paste0(var, ".hour")] +
dates_df[, paste0(var, ".minute")]
if ((unq_vals_n <- length(unique(dates_df[, paste0(var, ".day.minutes")]))) > 1) {
max_degree <- min(unq_vals_n, 5)
dates_df[, paste0(var, ".day.minutes.poly.", 1:max_degree)] <-
as.matrix(poly(dates_df[, paste0(var, ".day.minutes")], max_degree))
} else max_degree <- 0
# print(gp <- myplot_box(df=dates_df, ycol_names="PubDate.day.minutes",
# xcol_name=rsp_var))
# print(gp <- myplot_scatter(df=dates_df, xcol_name=".rownames",
# ycol_name="PubDate.day.minutes", colorcol_name=rsp_var))
# print(gp <- myplot_scatter(df=dates_df, xcol_name="PubDate.juliandate",
# ycol_name="PubDate.day.minutes.poly.1", colorcol_name=rsp_var))
# print(gp <- myplot_scatter(df=dates_df, xcol_name="PubDate.day.minutes",
# ycol_name="PubDate.day.minutes.poly.4", colorcol_name=rsp_var))
#
# print(gp <- myplot_scatter(df=dates_df, xcol_name="PubDate.juliandate",
# ycol_name="PubDate.day.minutes", colorcol_name=rsp_var, smooth=TRUE))
# print(gp <- myplot_scatter(df=dates_df, xcol_name="PubDate.juliandate",
# ycol_name="PubDate.day.minutes.poly.4", colorcol_name=rsp_var, smooth=TRUE))
# print(gp <- myplot_scatter(df=dates_df, xcol_name="PubDate.juliandate",
# ycol_name=c("PubDate.day.minutes", "PubDate.day.minutes.poly.4"),
# colorcol_name=rsp_var))
# print(gp <- myplot_scatter(df=subset(dates_df, Popular.fctr=="Y"),
# xcol_name=paste0(var, ".juliandate"),
# ycol_name=paste0(var, ".day.minutes", colorcol_name=rsp_var))
# print(gp <- myplot_box(df=dates_df, ycol_names=paste0(var, ".hour"),
# xcol_name=rsp_var))
# print(gp <- myplot_bar(df=dates_df, ycol_names=paste0(var, ".hour.fctr"),
# xcol_name=rsp_var,
# colorcol_name=paste0(var, ".hour.fctr")))
keep_feats <- paste(var,
c(".POSIX", ".year.fctr", ".month.fctr", ".date.fctr", ".wkday.fctr",
".wkend", ".hour.fctr", ".minute.fctr", ".second.fctr"), sep="")
if (max_degree > 0)
keep_feats <- union(keep_feats, paste(var,
paste0(".day.minutes.poly.", 1:max_degree), sep=""))
keep_feats <- intersect(keep_feats, names(dates_df))
}
#myprint_df(dates_df)
return(dates_df[, keep_feats])
}
if (!is.null(glb_date_vars)) {
glb_allobs_df <- cbind(glb_allobs_df,
myextract_dates_df(df=glb_allobs_df, vars=glb_date_vars,
id_vars=glb_id_var, rsp_var=glb_rsp_var))
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
paste(glb_date_vars, c("", ".POSIX"), sep=""))
for (feat in glb_date_vars) {
glb_allobs_df <- orderBy(reformulate(paste0(feat, ".POSIX")), glb_allobs_df)
# print(myplot_scatter(glb_allobs_df, xcol_name=paste0(feat, ".POSIX"),
# ycol_name=glb_rsp_var, colorcol_name=glb_rsp_var))
print(myplot_scatter(glb_allobs_df[glb_allobs_df[, paste0(feat, ".POSIX")] >=
strptime("2012-12-01", "%Y-%m-%d"), ],
xcol_name=paste0(feat, ".POSIX"),
ycol_name=glb_rsp_var, colorcol_name=paste0(feat, ".wkend")))
# Create features that measure the gap between previous timestamp in the data
require(zoo)
z <- zoo(as.numeric(as.POSIXlt(glb_allobs_df[, paste0(feat, ".POSIX")])))
glb_allobs_df[, paste0(feat, ".zoo")] <- z
print(head(glb_allobs_df[, c(glb_id_var, feat, paste0(feat, ".zoo"))]))
print(myplot_scatter(glb_allobs_df[glb_allobs_df[, paste0(feat, ".POSIX")] >
strptime("2012-10-01", "%Y-%m-%d"), ],
xcol_name=paste0(feat, ".zoo"), ycol_name=glb_rsp_var,
colorcol_name=glb_rsp_var))
b <- zoo(, seq(nrow(glb_allobs_df)))
last1 <- as.numeric(merge(z-lag(z, -1), b, all=TRUE)); last1[is.na(last1)] <- 0
glb_allobs_df[, paste0(feat, ".last1.log")] <- log(1 + last1)
print(gp <- myplot_box(df=glb_allobs_df[glb_allobs_df[,
paste0(feat, ".last1.log")] > 0, ],
ycol_names=paste0(feat, ".last1.log"),
xcol_name=glb_rsp_var))
last10 <- as.numeric(merge(z-lag(z, -10), b, all=TRUE)); last10[is.na(last10)] <- 0
glb_allobs_df[, paste0(feat, ".last10.log")] <- log(1 + last10)
print(gp <- myplot_box(df=glb_allobs_df[glb_allobs_df[,
paste0(feat, ".last10.log")] > 0, ],
ycol_names=paste0(feat, ".last10.log"),
xcol_name=glb_rsp_var))
last100 <- as.numeric(merge(z-lag(z, -100), b, all=TRUE)); last100[is.na(last100)] <- 0
glb_allobs_df[, paste0(feat, ".last100.log")] <- log(1 + last100)
print(gp <- myplot_box(df=glb_allobs_df[glb_allobs_df[,
paste0(feat, ".last100.log")] > 0, ],
ycol_names=paste0(feat, ".last100.log"),
xcol_name=glb_rsp_var))
glb_allobs_df <- orderBy(reformulate(glb_id_var), glb_allobs_df)
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
c(paste0(feat, ".zoo")))
# all2$last3 = as.numeric(merge(z-lag(z, -3), b, all = TRUE))
# all2$last5 = as.numeric(merge(z-lag(z, -5), b, all = TRUE))
# all2$last10 = as.numeric(merge(z-lag(z, -10), b, all = TRUE))
# all2$last20 = as.numeric(merge(z-lag(z, -20), b, all = TRUE))
# all2$last50 = as.numeric(merge(z-lag(z, -50), b, all = TRUE))
#
#
# # order table
# all2 = all2[order(all2$id),]
#
# ## fill in NAs
# # count averages
# na.avg = all2 %>% group_by(weekend, hour) %>% dplyr::summarise(
# last1=mean(last1, na.rm=TRUE),
# last3=mean(last3, na.rm=TRUE),
# last5=mean(last5, na.rm=TRUE),
# last10=mean(last10, na.rm=TRUE),
# last20=mean(last20, na.rm=TRUE),
# last50=mean(last50, na.rm=TRUE)
# )
#
# # fill in averages
# na.merge = merge(all2, na.avg, by=c("weekend","hour"))
# na.merge = na.merge[order(na.merge$id),]
# for(i in c("last1", "last3", "last5", "last10", "last20", "last50")) {
# y = paste0(i, ".y")
# idx = is.na(all2[[i]])
# all2[idx,][[i]] <- na.merge[idx,][[y]]
# }
# rm(na.avg, na.merge, b, i, idx, n, pd, sec, sh, y, z)
}
}
# check distribution of all numeric data
dsp_numeric_feats_dstrb <- function(feats_vctr) {
for (feat in feats_vctr) {
print(sprintf("feat: %s", feat))
if (glb_is_regression)
gp <- myplot_scatter(df=glb_allobs_df, ycol_name=glb_rsp_var, xcol_name=feat,
smooth=TRUE)
if (glb_is_classification)
gp <- myplot_box(df=glb_allobs_df, ycol_names=feat, xcol_name=glb_rsp_var)
if (inherits(glb_allobs_df[, feat], "factor"))
gp <- gp + facet_wrap(reformulate(feat))
print(gp)
}
}
# dsp_numeric_vars_dstrb(setdiff(names(glb_allobs_df),
# union(myfind_chr_cols_df(glb_allobs_df),
# c(glb_rsp_var_raw, glb_rsp_var))))
add_new_diag_feats <- function(obs_df, ref_df=glb_allobs_df) {
require(plyr)
obs_df <- mutate(obs_df,
# <col_name>.NA=is.na(<col_name>),
# <col_name>.fctr=factor(<col_name>,
# as.factor(union(obs_df$<col_name>, obs_twin_df$<col_name>))),
# <col_name>.fctr=relevel(factor(<col_name>,
# as.factor(union(obs_df$<col_name>, obs_twin_df$<col_name>))),
# "<ref_val>"),
# <col2_name>.fctr=relevel(factor(ifelse(<col1_name> == <val>, "<oth_val>", "<ref_val>")),
# as.factor(c("R", "<ref_val>")),
# ref="<ref_val>"),
# This doesn't work - use sapply instead
# <col_name>.fctr_num=grep(<col_name>, levels(<col_name>.fctr)),
#
# Date.my=as.Date(strptime(Date, "%m/%d/%y %H:%M")),
# Year=year(Date.my),
# Month=months(Date.my),
# Weekday=weekdays(Date.my)
# <col_name>=<table>[as.character(<col2_name>)],
# <col_name>=as.numeric(<col2_name>),
# <col_name> = trunc(<col2_name> / 100),
.rnorm = rnorm(n=nrow(obs_df))
)
# If levels of a factor are different across obs_df & glb_newobs_df; predict.glm fails
# Transformations not handled by mutate
# obs_df$<col_name>.fctr.num <- sapply(1:nrow(obs_df),
# function(row_ix) grep(obs_df[row_ix, "<col_name>"],
# levels(obs_df[row_ix, "<col_name>.fctr"])))
#print(summary(obs_df))
#print(sapply(names(obs_df), function(col) sum(is.na(obs_df[, col]))))
return(obs_df)
}
glb_allobs_df <- add_new_diag_feats(glb_allobs_df)## Loading required package: plyrrequire(dplyr)## Loading required package: dplyr
##
## Attaching package: 'dplyr'
##
## The following objects are masked from 'package:plyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarize
##
## The following object is masked from 'package:stats':
##
## filter
##
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union#stop(here"); sav_allobs_df <- glb_allobs_df # glb_allobs_df <- sav_allobs_df
# Merge some <descriptor>
# glb_allobs_df$<descriptor>.my <- glb_allobs_df$<descriptor>
# glb_allobs_df[grepl("\\bAIRPORT\\b", glb_allobs_df$<descriptor>.my),
# "<descriptor>.my"] <- "AIRPORT"
# glb_allobs_df$<descriptor>.my <-
# plyr::revalue(glb_allobs_df$<descriptor>.my, c(
# "ABANDONED BUILDING" = "OTHER",
# "##" = "##"
# ))
# print(<descriptor>_freq_df <- mycreate_sqlxtab_df(glb_allobs_df, c("<descriptor>.my")))
# # print(dplyr::filter(<descriptor>_freq_df, grepl("(MEDICAL|DENTAL|OFFICE)", <descriptor>.my)))
# # print(dplyr::filter(dplyr::select(glb_allobs_df, -<var.zoo>),
# # grepl("STORE", <descriptor>.my)))
# glb_exclude_vars_as_features <- c(glb_exclude_vars_as_features, "<descriptor>")
# Add logs of numerics that are not distributed normally -> do automatically ???
if (!is.null(glb_log_vars)) {
# Cycle thru glb_log_vars & create logs
# <col_name>.log=log(1 + <col.name>),
for (col in glb_log_vars)
glb_allobs_df[, paste0(col, ".log")] <- log(1 + glb_allobs_df[, col])
# Add raw_vars to glb_exclude_vars_as_features
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, glb_log_vars)
}
# Check distributions of newly transformed / extracted vars
# Enhancement: remove vars that were displayed ealier
dsp_numeric_feats_dstrb(feats_vctr=setdiff(names(glb_allobs_df),
c(myfind_chr_cols_df(glb_allobs_df), glb_rsp_var_raw, glb_rsp_var,
glb_exclude_vars_as_features)))## [1] "feat: PeopleInHousehold"
## [1] "feat: MetroAreaCode"## Warning in loop_apply(n, do.ply): Removed 34238 rows containing non-finite
## values (stat_boxplot).## Warning in loop_apply(n, do.ply): Removed 34238 rows containing missing
## values (stat_summary).
## [1] "feat: Age"
## [1] "feat: Hispanic"
## [1] "feat: CountryOfBirthCode"
## [1] "feat: .rnorm"
# Convert factors to dummy variables
# Build splines require(splines); bsBasis <- bs(training$age, df=3)
#pairs(subset(glb_trnobs_df, select=-c(col_symbol)))
# Check for glb_newobs_df & glb_trnobs_df features range mismatches
# Other diagnostics:
# print(subset(glb_trnobs_df, <col1_name> == max(glb_trnobs_df$<col1_name>, na.rm=TRUE) &
# <col2_name> <= mean(glb_trnobs_df$<col1_name>, na.rm=TRUE)))
# print(glb_trnobs_df[which.max(glb_trnobs_df$<col_name>),])
# print(<col_name>_freq_glb_trnobs_df <- mycreate_tbl_df(glb_trnobs_df, "<col_name>"))
# print(which.min(table(glb_trnobs_df$<col_name>)))
# print(which.max(table(glb_trnobs_df$<col_name>)))
# print(which.max(table(glb_trnobs_df$<col1_name>, glb_trnobs_df$<col2_name>)[, 2]))
# print(table(glb_trnobs_df$<col1_name>, glb_trnobs_df$<col2_name>))
# print(table(is.na(glb_trnobs_df$<col1_name>), glb_trnobs_df$<col2_name>))
# print(table(sign(glb_trnobs_df$<col1_name>), glb_trnobs_df$<col2_name>))
# print(mycreate_xtab_df(glb_trnobs_df, <col1_name>))
# print(mycreate_xtab_df(glb_trnobs_df, c(<col1_name>, <col2_name>)))
# print(<col1_name>_<col2_name>_xtab_glb_trnobs_df <-
# mycreate_xtab_df(glb_trnobs_df, c("<col1_name>", "<col2_name>")))
# <col1_name>_<col2_name>_xtab_glb_trnobs_df[is.na(<col1_name>_<col2_name>_xtab_glb_trnobs_df)] <- 0
# print(<col1_name>_<col2_name>_xtab_glb_trnobs_df <-
# mutate(<col1_name>_<col2_name>_xtab_glb_trnobs_df,
# <col3_name>=(<col1_name> * 1.0) / (<col1_name> + <col2_name>)))
# print(mycreate_sqlxtab_df(glb_allobs_df, c("<col1_name>", "<col2_name>")))
# print(<col2_name>_min_entity_arr <-
# sort(tapply(glb_trnobs_df$<col1_name>, glb_trnobs_df$<col2_name>, min, na.rm=TRUE)))
# print(<col1_name>_na_by_<col2_name>_arr <-
# sort(tapply(glb_trnobs_df$<col1_name>.NA, glb_trnobs_df$<col2_name>, mean, na.rm=TRUE)))
# Other plots:
# print(myplot_box(df=glb_trnobs_df, ycol_names="<col1_name>"))
# print(myplot_box(df=glb_trnobs_df, ycol_names="<col1_name>", xcol_name="<col2_name>"))
# print(myplot_line(subset(glb_trnobs_df, Symbol %in% c("CocaCola", "ProcterGamble")),
# "Date.POSIX", "StockPrice", facet_row_colnames="Symbol") +
# geom_vline(xintercept=as.numeric(as.POSIXlt("2003-03-01"))) +
# geom_vline(xintercept=as.numeric(as.POSIXlt("1983-01-01")))
# )
# print(myplot_line(subset(glb_trnobs_df, Date.POSIX > as.POSIXct("2004-01-01")),
# "Date.POSIX", "StockPrice") +
# geom_line(aes(color=Symbol)) +
# coord_cartesian(xlim=c(as.POSIXct("1990-01-01"),
# as.POSIXct("2000-01-01"))) +
# coord_cartesian(ylim=c(0, 250)) +
# geom_vline(xintercept=as.numeric(as.POSIXlt("1997-09-01"))) +
# geom_vline(xintercept=as.numeric(as.POSIXlt("1997-11-01")))
# )
# print(myplot_scatter(glb_allobs_df, "<col1_name>", "<col2_name>", smooth=TRUE))
# print(myplot_scatter(glb_allobs_df, "<col1_name>", "<col2_name>", colorcol_name="<Pred.fctr>") +
# geom_point(data=subset(glb_allobs_df, <condition>),
# mapping=aes(x=<x_var>, y=<y_var>), color="red", shape=4, size=5))
rm(srt_allobs_df, last1, last10, last100, pd)## Warning in rm(srt_allobs_df, last1, last10, last100, pd): object
## 'srt_allobs_df' not found## Warning in rm(srt_allobs_df, last1, last10, last100, pd): object 'last1'
## not found## Warning in rm(srt_allobs_df, last1, last10, last100, pd): object 'last10'
## not found## Warning in rm(srt_allobs_df, last1, last10, last100, pd): object 'last100'
## not found## Warning in rm(srt_allobs_df, last1, last10, last100, pd): object 'pd' not
## foundglb_chunks_df <- myadd_chunk(glb_chunks_df, "scrub.data", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 2 inspect.data 2 0 19.517 42.418 22.901
## 3 scrub.data 2 1 42.419 NA NAStep 2.1: scrub data
# Options:
# 1. Not fill missing vars
# 2. Fill missing numerics with a different algorithm
# 3. Fill missing chars with data based on clusters
mycheck_problem_data(glb_allobs_df)## [1] "numeric data missing in glb_allobs_df: "
## MetroAreaCode EmploymentStatus.fctr
## 34238 25789
## [1] "numeric data w/ 0s in glb_allobs_df: "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in glb_allobs_df: "
## named integer(0)
## [1] "numeric data w/ NaNs in glb_allobs_df: "
## named integer(0)
## [1] "string data missing in glb_allobs_df: "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames
## NA 0# if (!is.null(glb_force_0_to_NA_vars)) {
# for (feat in glb_force_0_to_NA_vars) {
# warning("Forcing ", sum(glb_allobs_df[, feat] == 0),
# " obs with ", feat, " 0s to NAs")
# glb_allobs_df[glb_allobs_df[, feat] == 0, feat] <- NA
# }
# }
mycheck_problem_data(glb_allobs_df)## [1] "numeric data missing in glb_allobs_df: "
## MetroAreaCode EmploymentStatus.fctr
## 34238 25789
## [1] "numeric data w/ 0s in glb_allobs_df: "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in glb_allobs_df: "
## named integer(0)
## [1] "numeric data w/ NaNs in glb_allobs_df: "
## named integer(0)
## [1] "string data missing in glb_allobs_df: "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames
## NA 0dsp_catgs <- function() {
print("NewsDesk:")
print(table(glb_allobs_df$NewsDesk))
print("SectionName:")
print(table(glb_allobs_df$SectionName))
print("SubsectionName:")
print(table(glb_allobs_df$SubsectionName))
}
# sel_obs <- function(Popular=NULL,
# NewsDesk=NULL, SectionName=NULL, SubsectionName=NULL,
# Headline.contains=NULL, Snippet.contains=NULL, Abstract.contains=NULL,
# Headline.pfx=NULL, NewsDesk.nb=NULL, .clusterid=NULL, myCategory=NULL,
# perl=FALSE) {
sel_obs <- function(vars_lst) {
tmp_df <- glb_allobs_df
# Does not work for Popular == NAs ???
if (!is.null(Popular)) {
if (is.na(Popular))
tmp_df <- tmp_df[is.na(tmp_df$Popular), ] else
tmp_df <- tmp_df[tmp_df$Popular == Popular, ]
}
if (!is.null(NewsDesk))
tmp_df <- tmp_df[tmp_df$NewsDesk == NewsDesk, ]
if (!is.null(SectionName))
tmp_df <- tmp_df[tmp_df$SectionName == SectionName, ]
if (!is.null(SubsectionName))
tmp_df <- tmp_df[tmp_df$SubsectionName == SubsectionName, ]
if (!is.null(Headline.contains))
tmp_df <-
tmp_df[grep(Headline.contains, tmp_df$Headline, perl=perl), ]
if (!is.null(Snippet.contains))
tmp_df <-
tmp_df[grep(Snippet.contains, tmp_df$Snippet, perl=perl), ]
if (!is.null(Abstract.contains))
tmp_df <-
tmp_df[grep(Abstract.contains, tmp_df$Abstract, perl=perl), ]
if (!is.null(Headline.pfx)) {
if (length(grep("Headline.pfx", names(tmp_df), fixed=TRUE, value=TRUE))
> 0) tmp_df <-
tmp_df[tmp_df$Headline.pfx == Headline.pfx, ] else
warning("glb_allobs_df does not contain Headline.pfx; ignoring that filter")
}
if (!is.null(NewsDesk.nb)) {
if (any(grepl("NewsDesk.nb", names(tmp_df), fixed=TRUE)) > 0)
tmp_df <-
tmp_df[tmp_df$NewsDesk.nb == NewsDesk.nb, ] else
warning("glb_allobs_df does not contain NewsDesk.nb; ignoring that filter")
}
if (!is.null(.clusterid)) {
if (any(grepl(".clusterid", names(tmp_df), fixed=TRUE)) > 0)
tmp_df <-
tmp_df[tmp_df$clusterid == clusterid, ] else
warning("glb_allobs_df does not contain clusterid; ignoring that filter") }
if (!is.null(myCategory)) {
if (!(myCategory %in% names(glb_allobs_df)))
tmp_df <-
tmp_df[tmp_df$myCategory == myCategory, ] else
warning("glb_allobs_df does not contain myCategory; ignoring that filter")
}
return(glb_allobs_df$UniqueID %in% tmp_df$UniqueID)
}
dsp_obs <- function(..., cols=c(NULL), all=FALSE) {
tmp_df <- glb_allobs_df[sel_obs(...),
union(c("UniqueID", "Popular", "myCategory", "Headline"), cols), FALSE]
if(all) { print(tmp_df) } else { myprint_df(tmp_df) }
}
#dsp_obs(Popular=1, NewsDesk="", SectionName="", Headline.contains="Boehner")
# dsp_obs(Popular=1, NewsDesk="", SectionName="")
# dsp_obs(Popular=NA, NewsDesk="", SectionName="")
dsp_tbl <- function(...) {
tmp_entity_df <- glb_allobs_df[sel_obs(...), ]
tmp_tbl <- table(tmp_entity_df$NewsDesk,
tmp_entity_df$SectionName,
tmp_entity_df$SubsectionName,
tmp_entity_df$Popular, useNA="ifany")
#print(names(tmp_tbl))
#print(dimnames(tmp_tbl))
print(tmp_tbl)
}
dsp_hdlxtab <- function(str)
print(mycreate_sqlxtab_df(glb_allobs_df[sel_obs(Headline.contains=str), ],
c("Headline.pfx", "Headline", glb_rsp_var)))
#dsp_hdlxtab("(1914)|(1939)")
dsp_catxtab <- function(str)
print(mycreate_sqlxtab_df(glb_allobs_df[sel_obs(Headline.contains=str), ],
c("Headline.pfx", "NewsDesk", "SectionName", "SubsectionName", glb_rsp_var)))
# dsp_catxtab("1914)|(1939)")
# dsp_catxtab("19(14|39|64):")
# dsp_catxtab("19..:")
# Create myCategory <- NewsDesk#SectionName#SubsectionName
# Fix some data before merging categories
# glb_allobs_df[sel_obs(Headline.contains="Your Turn:", NewsDesk=""),
# "NewsDesk"] <- "Styles"
# glb_allobs_df[sel_obs(Headline.contains="School", NewsDesk="", SectionName="U.S.",
# SubsectionName=""),
# "SubsectionName"] <- "Education"
# glb_allobs_df[sel_obs(Headline.contains="Today in Small Business:", NewsDesk="Business"),
# "SectionName"] <- "Business Day"
# glb_allobs_df[sel_obs(Headline.contains="Today in Small Business:", NewsDesk="Business"),
# "SubsectionName"] <- "Small Business"
# glb_allobs_df[sel_obs(Headline.contains="Readers Respond:"),
# "SectionName"] <- "Opinion"
# glb_allobs_df[sel_obs(Headline.contains="Readers Respond:"),
# "SubsectionName"] <- "Room For Debate"
# glb_allobs_df[sel_obs(NewsDesk="Business", SectionName="", SubsectionName="", Popular=NA),
# "SubsectionName"] <- "Small Business"
# print(glb_allobs_df[glb_allobs_df$UniqueID %in% c(7973),
# c("UniqueID", "Headline", "myCategory", "NewsDesk", "SectionName", "SubsectionName")])
#
# glb_allobs_df[sel_obs(NewsDesk="Business", SectionName="", SubsectionName=""),
# "SectionName"] <- "Technology"
# print(glb_allobs_df[glb_allobs_df$UniqueID %in% c(5076, 5736, 5924, 5911, 6532),
# c("UniqueID", "Headline", "myCategory", "NewsDesk", "SectionName", "SubsectionName")])
#
# glb_allobs_df[sel_obs(SectionName="Health"),
# "NewsDesk"] <- "Science"
# glb_allobs_df[sel_obs(SectionName="Travel"),
# "NewsDesk"] <- "Travel"
#
# glb_allobs_df[sel_obs(SubsectionName="Fashion & Style"),
# "SectionName"] <- ""
# glb_allobs_df[sel_obs(SubsectionName="Fashion & Style"),
# "SubsectionName"] <- ""
# glb_allobs_df[sel_obs(NewsDesk="Styles", SectionName="", SubsectionName="", Popular=1),
# "SectionName"] <- "U.S."
# print(glb_allobs_df[glb_allobs_df$UniqueID %in% c(5486),
# c("UniqueID", "Headline", "myCategory", "NewsDesk", "SectionName", "SubsectionName")])
#
# glb_allobs_df$myCategory <- paste(glb_allobs_df$NewsDesk,
# glb_allobs_df$SectionName,
# glb_allobs_df$SubsectionName,
# sep="#")
# dsp_obs( Headline.contains="Music:"
# #,NewsDesk=""
# #,SectionName=""
# #,SubsectionName="Fashion & Style"
# #,Popular=1 #NA
# ,cols= c("UniqueID", "Headline", "Popular", "myCategory",
# "NewsDesk", "SectionName", "SubsectionName"),
# all=TRUE)
# dsp_obs( Headline.contains="."
# ,NewsDesk=""
# ,SectionName="Opinion"
# ,SubsectionName=""
# #,Popular=1 #NA
# ,cols= c("UniqueID", "Headline", "Popular", "myCategory",
# "NewsDesk", "SectionName", "SubsectionName"),
# all=TRUE)
# Merge some categories
# glb_allobs_df$myCategory <-
# plyr::revalue(glb_allobs_df$myCategory, c(
# "#Business Day#Dealbook" = "Business#Business Day#Dealbook",
# "#Business Day#Small Business" = "Business#Business Day#Small Business",
# "#Crosswords/Games#" = "Business#Crosswords/Games#",
# "Business##" = "Business#Technology#",
# "#Open#" = "Business#Technology#",
# "#Technology#" = "Business#Technology#",
#
# "#Arts#" = "Culture#Arts#",
# "Culture##" = "Culture#Arts#",
#
# "#World#Asia Pacific" = "Foreign#World#Asia Pacific",
# "Foreign##" = "Foreign#World#",
#
# "#N.Y. / Region#" = "Metro#N.Y. / Region#",
#
# "#Opinion#" = "OpEd#Opinion#",
# "OpEd##" = "OpEd#Opinion#",
#
# "#Health#" = "Science#Health#",
# "Science##" = "Science#Health#",
#
# "Styles##" = "Styles##Fashion",
# "Styles#Health#" = "Science#Health#",
# "Styles#Style#Fashion & Style" = "Styles##Fashion",
#
# "#Travel#" = "Travel#Travel#",
#
# "Magazine#Magazine#" = "myOther",
# "National##" = "myOther",
# "National#U.S.#Politics" = "myOther",
# "Sports##" = "myOther",
# "Sports#Sports#" = "myOther",
# "#U.S.#" = "myOther",
#
#
# # "Business##Small Business" = "Business#Business Day#Small Business",
# #
# # "#Opinion#" = "#Opinion#Room For Debate",
# "##" = "##"
# # "Business##" = "Business#Business Day#Dealbook",
# # "Foreign#World#" = "Foreign##",
# # "#Open#" = "Other",
# # "#Opinion#The Public Editor" = "OpEd#Opinion#",
# # "Styles#Health#" = "Styles##",
# # "Styles#Style#Fashion & Style" = "Styles##",
# # "#U.S.#" = "#U.S.#Education",
# ))
# ctgry_xtab_df <- orderBy(reformulate(c("-", ".n")),
# mycreate_sqlxtab_df(glb_allobs_df,
# c("myCategory", "NewsDesk", "SectionName", "SubsectionName", glb_rsp_var)))
# myprint_df(ctgry_xtab_df)
# write.table(ctgry_xtab_df, paste0(glb_out_pfx, "ctgry_xtab.csv"),
# row.names=FALSE)
# ctgry_cast_df <- orderBy(~ -Y -NA, dcast(ctgry_xtab_df,
# myCategory + NewsDesk + SectionName + SubsectionName ~
# Popular.fctr, sum, value.var=".n"))
# myprint_df(ctgry_cast_df)
# write.table(ctgry_cast_df, paste0(glb_out_pfx, "ctgry_cast.csv"),
# row.names=FALSE)
# print(ctgry_sum_tbl <- table(glb_allobs_df$myCategory, glb_allobs_df[, glb_rsp_var],
# useNA="ifany"))
dsp_chisq.test <- function(...) {
sel_df <- glb_allobs_df[sel_obs(...) &
!is.na(glb_allobs_df$Popular), ]
sel_df$.marker <- 1
ref_df <- glb_allobs_df[!is.na(glb_allobs_df$Popular), ]
mrg_df <- merge(ref_df[, c(glb_id_var, "Popular")],
sel_df[, c(glb_id_var, ".marker")], all.x=TRUE)
mrg_df[is.na(mrg_df)] <- 0
print(mrg_tbl <- table(mrg_df$.marker, mrg_df$Popular))
print("Rows:Selected; Cols:Popular")
#print(mrg_tbl)
print(chisq.test(mrg_tbl))
}
# dsp_chisq.test(Headline.contains="[Ee]bola")
# dsp_chisq.test(Snippet.contains="[Ee]bola")
# dsp_chisq.test(Abstract.contains="[Ee]bola")
# print(mycreate_sqlxtab_df(glb_allobs_df[sel_obs(Headline.contains="[Ee]bola"), ],
# c(glb_rsp_var, "NewsDesk", "SectionName", "SubsectionName")))
# print(table(glb_allobs_df$NewsDesk, glb_allobs_df$SectionName))
# print(table(glb_allobs_df$SectionName, glb_allobs_df$SubsectionName))
# print(table(glb_allobs_df$NewsDesk, glb_allobs_df$SectionName, glb_allobs_df$SubsectionName))
# glb_allobs_df$myCategory.fctr <- as.factor(glb_allobs_df$myCategory)
# glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
# c("myCategory", "NewsDesk", "SectionName", "SubsectionName"))
# Copy Headline into Snipper & Abstract if they are empty
# print(glb_allobs_df[nchar(glb_allobs_df[, "Snippet"]) == 0, c("Headline", "Snippet")])
# print(glb_allobs_df[glb_allobs_df$Headline == glb_allobs_df$Snippet,
# c("UniqueID", "Headline", "Snippet")])
# glb_allobs_df[nchar(glb_allobs_df[, "Snippet"]) == 0, "Snippet"] <-
# glb_allobs_df[nchar(glb_allobs_df[, "Snippet"]) == 0, "Headline"]
#
# print(glb_allobs_df[nchar(glb_allobs_df[, "Abstract"]) == 0, c("Headline", "Abstract")])
# print(glb_allobs_df[glb_allobs_df$Headline == glb_allobs_df$Abstract,
# c("UniqueID", "Headline", "Abstract")])
# glb_allobs_df[nchar(glb_allobs_df[, "Abstract"]) == 0, "Abstract"] <-
# glb_allobs_df[nchar(glb_allobs_df[, "Abstract"]) == 0, "Headline"]
# WordCount_0_df <- subset(glb_allobs_df, WordCount == 0)
# table(WordCount_0_df$Popular, WordCount_0_df$WordCount, useNA="ifany")
# myprint_df(WordCount_0_df[,
# c("UniqueID", "Popular", "WordCount", "Headline")])Step 2.1: scrub data
glb_chunks_df <- myadd_chunk(glb_chunks_df, "encode.data", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 3 scrub.data 2 1 42.419 45.384 2.965
## 4 encode.data 2 2 45.384 NA NA# map_<col_name>_df <- myimport_data(
# url="<map_url>",
# comment="map_<col_name>_df", print_diagn=TRUE)
# map_<col_name>_df <- read.csv(paste0(getwd(), "/data/<file_name>.csv"), strip.white=TRUE)
# glb_trnobs_df <- mymap_codes(glb_trnobs_df, "<from_col_name>", "<to_col_name>",
# map_<to_col_name>_df, map_join_col_name="<map_join_col_name>",
# map_tgt_col_name="<to_col_name>")
# glb_newobs_df <- mymap_codes(glb_newobs_df, "<from_col_name>", "<to_col_name>",
# map_<to_col_name>_df, map_join_col_name="<map_join_col_name>",
# map_tgt_col_name="<to_col_name>")
# glb_trnobs_df$<col_name>.fctr <- factor(glb_trnobs_df$<col_name>,
# as.factor(union(glb_trnobs_df$<col_name>, glb_newobs_df$<col_name>)))
# glb_newobs_df$<col_name>.fctr <- factor(glb_newobs_df$<col_name>,
# as.factor(union(glb_trnobs_df$<col_name>, glb_newobs_df$<col_name>)))
#sav_allobs_df <- glb_allobs_df; glb_allobs_df <- sav_allobs_df
if (!is.null(glb_map_vars)) {
for (feat in glb_map_vars) {
map_df <- myimport_data(url=glb_map_urls[[feat]],
comment="map_df",
print_diagn=TRUE)
glb_allobs_df <- mymap_codes(glb_allobs_df, feat, names(map_df)[2],
map_df, map_join_col_name=names(map_df)[1],
map_tgt_col_name=names(map_df)[2])
}
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, glb_map_vars)
}## [1] "Reading file ./data/MetroAreaCodes.csv..."
## [1] "dimensions of data in ./data/MetroAreaCodes.csv: 271 rows x 2 cols"
## Code MetroArea
## 1 460 Appleton-Oshkosh-Neenah, WI
## 2 3000 Grand Rapids-Muskegon-Holland, MI
## 3 3160 Greenville-Spartanburg-Anderson, SC
## 4 3610 Jamestown, NY
## 5 3720 Kalamazoo-Battle Creek, MI
## 6 6450 Portsmouth-Rochester, NH-ME
## Code MetroArea
## 23 12260 Augusta-Richmond County, GA-SC
## 83 22420 Flint, MI
## 116 27900 Joplin, MO
## 172 36540 Omaha-Council Bluffs, NE-IA
## 195 40140 Riverside-San Bernardino, CA
## 262 73450 Hartford-West Hartford-East Hartford, CT
## Code MetroArea
## 266 76750 Portland-South Portland, ME
## 267 77200 Providence-Fall River-Warwick, MA-RI
## 268 77350 Rochester-Dover, NH-ME
## 269 78100 Springfield, MA-CT
## 270 78700 Waterbury, CT
## 271 79600 Worcester, MA-CT
## 'data.frame': 271 obs. of 2 variables:
## $ Code : int 460 3000 3160 3610 3720 6450 10420 10500 10580 10740 ...
## $ MetroArea: chr "Appleton-Oshkosh-Neenah, WI" "Grand Rapids-Muskegon-Holland, MI" "Greenville-Spartanburg-Anderson, SC" "Jamestown, NY" ...
## - attr(*, "comment")= chr "map_df"
## NULL
## MetroAreaCode MetroArea .n
## 1 NA <NA> 34238
## 2 35620 New York-Northern New Jersey-Long Island, NY-NJ-PA 5409
## 3 47900 Washington-Arlington-Alexandria, DC-VA-MD-WV 4177
## 4 31100 Los Angeles-Long Beach-Santa Ana, CA 4102
## 5 37980 Philadelphia-Camden-Wilmington, PA-NJ-DE 2855
## 6 16980 Chicago-Naperville-Joliet, IN-IN-WI 2772
## MetroAreaCode MetroArea .n
## 4 31100 Los Angeles-Long Beach-Santa Ana, CA 4102
## 23 38060 Phoenix-Mesa-Scottsdale, AZ 971
## 73 24340 Grand Rapids-Wyoming, MI 304
## 92 27140 Jackson, MS 222
## 208 15180 Brownsville-Harlingen, TX 79
## 214 33140 Michigan City-La Porte, IN 77
## MetroAreaCode MetroArea .n
## 260 46660 Valdosta, GA 42
## 261 47580 Warner Robins, GA 42
## 262 14060 Bloomington-Normal IL 40
## 263 44220 Springfield, OH 34
## 264 36140 Ocean City, NJ 30
## 265 14540 Bowling Green, KY 29## Warning in loop_apply(n, do.ply): Removed 1 rows containing missing values
## (position_stack).## Warning in loop_apply(n, do.ply): position_stack requires constant width:
## output may be incorrect
## [1] "Reading file ./data/CountryCodes.csv..."
## [1] "dimensions of data in ./data/CountryCodes.csv: 149 rows x 2 cols"
## Code Country
## 1 57 United States
## 2 66 Guam
## 3 73 Puerto Rico
## 4 78 U. S. Virgin Islands
## 5 96 Other U. S. Island Areas
## 6 100 Albania
## Code Country
## 42 158 Armenia
## 56 207 China
## 64 215 Japan
## 97 323 Bahamas
## 111 361 Bolivia
## 115 365 Ecuador
## Code Country
## 144 508 Fiji
## 145 515 New Zealand
## 146 523 Tonga
## 147 527 Samoa
## 148 528 Oceania, not specified
## 149 555 Elsewhere
## 'data.frame': 149 obs. of 2 variables:
## $ Code : int 57 66 73 78 96 100 102 103 104 105 ...
## $ Country: chr "United States" "Guam" "Puerto Rico" "U. S. Virgin Islands" ...
## - attr(*, "comment")= chr "map_df"
## NULL
## CountryOfBirthCode Country .n
## 1 57 United States 115063
## 2 303 Mexico 3921
## 3 233 Philippines 839
## 4 210 India 770
## 5 207 China 581
## 6 73 Puerto Rico 518
## CountryOfBirthCode Country .n
## 18 215 Japan 187
## 20 163 Russia 173
## 86 160 Belarus 24
## 101 78 U. S. Virgin Islands 17
## 138 149 Slovakia 6
## 157 142 Northern Ireland 2
## CountryOfBirthCode Country .n
## 156 425 <NA> 3
## 157 142 Northern Ireland 2
## 158 228 <NA> 2
## 159 453 Tanzania 2
## 160 430 <NA> 1
## 161 460 <NA> 1## Warning in loop_apply(n, do.ply): Removed 17 rows containing missing values
## (position_stack).## Warning in loop_apply(n, do.ply): position_stack requires constant width:
## output may be incorrect
#stop(here"); sav_allobs_df <- glb_allobs_df; glb_allobs_df <- sav_allobs_df
for (feat in glb_assign_vars) {
new_feat <- paste0(feat, ".my")
print(sprintf("Forced Assignments for: %s -> %s...", feat, new_feat))
glb_allobs_df[, new_feat] <- glb_allobs_df[, feat]
pairs <- glb_assign_pairs_lst[[feat]]
for (pair_ix in 1:length(pairs$from)) {
if (is.na(pairs$from[pair_ix]))
nobs <- nrow(filter(glb_allobs_df,
is.na(eval(parse(text=feat),
envir=glb_allobs_df)))) else
nobs <- sum(glb_allobs_df[, feat] == pairs$from[pair_ix])
#nobs <- nrow(filter(glb_allobs_df, is.na(Married.fctr))) ; print(nobs)
if ((is.na(pairs$from[pair_ix])) && (is.na(pairs$to[pair_ix])))
stop("what are you trying to do ???")
if (is.na(pairs$from[pair_ix]))
glb_allobs_df[is.na(glb_allobs_df[, feat]), new_feat] <-
pairs$to[pair_ix] else
glb_allobs_df[glb_allobs_df[, feat] == pairs$from[pair_ix], new_feat] <-
pairs$to[pair_ix]
print(sprintf(" %s -> %s for %s obs",
pairs$from[pair_ix], pairs$to[pair_ix], format(nobs, big.mark=",")))
}
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, glb_assign_vars)
}## [1] "Forced Assignments for: Married -> Married.my..."
## [1] " NA -> NA.my for 25,338 obs"
## [1] "Forced Assignments for: Education -> Education.my..."
## [1] " NA -> NA.my for 25,338 obs"
## [1] "Forced Assignments for: Industry -> Industry.my..."
## [1] " NA -> NA.my for 65,060 obs"
## [1] "Forced Assignments for: MetroArea -> MetroArea.my..."
## [1] " NA -> NA.my for 34,238 obs"
## [1] "Forced Assignments for: Country -> Country.my..."
## [1] " NA -> NA.my for 176 obs"Step 2.2: encode data
glb_chunks_df <- myadd_chunk(glb_chunks_df, "manage.missing.data", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 4 encode.data 2 2 45.384 63.659 18.275
## 5 manage.missing.data 2 3 63.659 NA NA# print(sapply(names(glb_trnobs_df), function(col) sum(is.na(glb_trnobs_df[, col]))))
# print(sapply(names(glb_newobs_df), function(col) sum(is.na(glb_newobs_df[, col]))))
# glb_trnobs_df <- na.omit(glb_trnobs_df)
# glb_newobs_df <- na.omit(glb_newobs_df)
# df[is.na(df)] <- 0
mycheck_problem_data(glb_allobs_df)## [1] "numeric data missing in : "
## MetroAreaCode EmploymentStatus.fctr
## 34238 25789
## [1] "numeric data w/ 0s in : "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in : "
## named integer(0)
## [1] "numeric data w/ NaNs in : "
## named integer(0)
## [1] "string data missing in : "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames MetroArea Country
## NA 0 NA NA
## Married.my Education.my Industry.my MetroArea.my
## 0 0 0 0
## Country.my
## 0# Not refactored into mydsutils.R since glb_*_df might be reassigned
glb_impute_missing_data <- function() {
require(mice)
set.seed(glb_mice_complete.seed)
inp_impent_df <- glb_allobs_df[, setdiff(names(glb_allobs_df),
union(glb_exclude_vars_as_features, glb_rsp_var))]
print("Summary before imputation: ")
print(summary(inp_impent_df))
out_impent_df <- complete(mice(inp_impent_df))
print(summary(out_impent_df))
# complete(mice()) changes attributes of factors even though values don't change
ret_vars <- sapply(names(out_impent_df),
function(col) ifelse(!identical(out_impent_df[, col], inp_impent_df[, col]),
col, ""))
ret_vars <- ret_vars[ret_vars != ""]
return(out_impent_df[, ret_vars])
}
if (glb_impute_na_data &&
(length(myfind_numerics_missing(glb_allobs_df)) > 0) &&
(ncol(nonna_df <- glb_impute_missing_data()) > 0)) {
for (col in names(nonna_df)) {
glb_allobs_df[, paste0(col, ".nonNA")] <- nonna_df[, col]
glb_exclude_vars_as_features <- c(glb_exclude_vars_as_features, col)
}
}
mycheck_problem_data(glb_allobs_df, terminate = TRUE)## [1] "numeric data missing in : "
## MetroAreaCode EmploymentStatus.fctr
## 34238 25789
## [1] "numeric data w/ 0s in : "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in : "
## named integer(0)
## [1] "numeric data w/ NaNs in : "
## named integer(0)
## [1] "string data missing in : "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames MetroArea Country
## NA 0 NA NA
## Married.my Education.my Industry.my MetroArea.my
## 0 0 0 0
## Country.my
## 0Step 2.3: manage missing data
#```{r extract_features, cache=FALSE, eval=!is.null(glb_txt_vars)}
glb_chunks_df <- myadd_chunk(glb_chunks_df, "extract.features", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 5 manage.missing.data 2 3 63.659 64.136 0.478
## 6 extract.features 3 0 64.137 NA NAextract.features_chunk_df <- myadd_chunk(NULL, "extract.features_bgn")## label step_major step_minor bgn end elapsed
## 1 extract.features_bgn 1 0 64.144 NA NA# Options:
# Select Tf, log(1 + Tf), Tf-IDF or BM25Tf-IDf
# Create new features that help prediction
# <col_name>.lag.2 <- lag(zoo(glb_trnobs_df$<col_name>), -2, na.pad=TRUE)
# glb_trnobs_df[, "<col_name>.lag.2"] <- coredata(<col_name>.lag.2)
# <col_name>.lag.2 <- lag(zoo(glb_newobs_df$<col_name>), -2, na.pad=TRUE)
# glb_newobs_df[, "<col_name>.lag.2"] <- coredata(<col_name>.lag.2)
#
# glb_newobs_df[1, "<col_name>.lag.2"] <- glb_trnobs_df[nrow(glb_trnobs_df) - 1,
# "<col_name>"]
# glb_newobs_df[2, "<col_name>.lag.2"] <- glb_trnobs_df[nrow(glb_trnobs_df),
# "<col_name>"]
# glb_allobs_df <- mutate(glb_allobs_df,
# A.P.http=ifelse(grepl("http",Added,fixed=TRUE), 1, 0)
# )
#
# glb_trnobs_df <- mutate(glb_trnobs_df,
# )
#
# glb_newobs_df <- mutate(glb_newobs_df,
# )
# Create factors of string variables
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "factorize.str.vars"), major.inc=TRUE)## label step_major step_minor bgn end
## 1 extract.features_bgn 1 0 64.144 64.154
## 2 extract.features_factorize.str.vars 2 0 64.155 NA
## elapsed
## 1 0.01
## 2 NA#stop(here"); sav_allobs_df <- glb_allobs_df; #glb_allobs_df <- sav_allobs_df
print(str_vars <- myfind_chr_cols_df(glb_allobs_df))## Region State Married
## "Region" "State" "Married"
## Sex Education Race
## "Sex" "Education" "Race"
## Citizenship EmploymentStatus Industry
## "Citizenship" "EmploymentStatus" "Industry"
## .rownames .src MetroArea
## ".rownames" ".src" "MetroArea"
## Country Married.my Education.my
## "Country" "Married.my" "Education.my"
## Industry.my MetroArea.my Country.my
## "Industry.my" "MetroArea.my" "Country.my"if (length(str_vars <- setdiff(str_vars,
glb_exclude_vars_as_features)) > 0) {
for (var in str_vars) {
warning("Creating factors of string variable: ", var,
": # of unique values: ", length(unique(glb_allobs_df[, var])))
glb_allobs_df[, paste0(var, ".fctr")] <- factor(glb_allobs_df[, var],
as.factor(unique(glb_allobs_df[, var])))
# glb_trnobs_df[, paste0(var, ".fctr")] <- factor(glb_trnobs_df[, var],
# as.factor(unique(glb_allobs_df[, var])))
# glb_newobs_df[, paste0(var, ".fctr")] <- factor(glb_newobs_df[, var],
# as.factor(unique(glb_allobs_df[, var])))
}
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features, str_vars)
}## Warning: Creating factors of string variable: Region: # of unique values: 4## Warning: Creating factors of string variable: State: # of unique values: 51## Warning: Creating factors of string variable: Sex: # of unique values: 2## Warning: Creating factors of string variable: Race: # of unique values: 6## Warning: Creating factors of string variable: Citizenship: # of unique
## values: 3## Warning: Creating factors of string variable: Married.my: # of unique
## values: 6## Warning: Creating factors of string variable: Education.my: # of unique
## values: 9## Warning: Creating factors of string variable: Industry.my: # of unique
## values: 15## Warning: Creating factors of string variable: MetroArea.my: # of unique
## values: 265## Warning: Creating factors of string variable: Country.my: # of unique
## values: 145if (!is.null(glb_txt_vars)) {
require(foreach)
require(gsubfn)
require(stringr)
require(tm)
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "process.text"), major.inc=TRUE)
chk_pattern_freq <- function(re_str, ignore.case=TRUE) {
match_mtrx <- str_extract_all(txt_vctr, regex(re_str, ignore_case=ignore.case),
simplify=TRUE)
match_df <- as.data.frame(match_mtrx[match_mtrx != ""])
names(match_df) <- "pattern"
return(mycreate_sqlxtab_df(match_df, "pattern"))
}
#tmp_freq_df <- chk_pattern_freq("\\bNew (\\w)+", ignore.case=FALSE)
#subset(chk_pattern_freq("\\bNew (\\w)+", ignore.case=FALSE), grepl("New [[:upper:]]", pattern))
#chk_pattern_freq("\\bnew (\\W)+")
chk_subfn <- function(pos_ix) {
re_str <- gsubfn_args_lst[["re_str"]][[pos_ix]]
print("re_str:"); print(re_str)
rp_frmla <- gsubfn_args_lst[["rp_frmla"]][[pos_ix]]
print("rp_frmla:"); print(rp_frmla, showEnv=FALSE)
tmp_vctr <- grep(re_str, txt_vctr, value=TRUE, ignore.case=TRUE)[1:5]
print("Before:")
print(tmp_vctr)
print("After:")
print(gsubfn(re_str, rp_frmla, tmp_vctr, ignore.case=TRUE))
}
#chk_subfn(1)
myapply_gsub <- function(...) {
if ((length_lst <- length(names(gsub_map_lst))) == 0)
return(txt_vctr)
for (ptn_ix in 1:length_lst) {
print(sprintf("running gsub for %02d (of %02d): #%s#...", ptn_ix,
length(names(gsub_map_lst)), names(gsub_map_lst)[ptn_ix]))
txt_vctr <- gsub(names(gsub_map_lst)[ptn_ix], gsub_map_lst[[ptn_ix]],
txt_vctr, ...)
}
return(txt_vctr)
}
myapply_txtmap <- function(txt_vctr, ...) {
nrows <- nrow(glb_txt_map_df)
for (ptn_ix in 1:nrows) {
print(sprintf("running gsub for %02d (of %02d): #%s#...", ptn_ix,
nrows, glb_txt_map_df[ptn_ix, "rex_str"]))
txt_vctr <- gsub(glb_txt_map_df[ptn_ix, "rex_str"],
glb_txt_map_df[ptn_ix, "rpl_str"],
txt_vctr, ...)
}
return(txt_vctr)
}
chk.equal <- function(bgn, end) {
print(all.equal(sav_txt_lst[["Headline"]][bgn:end], glb_txt_lst[["Headline"]][bgn:end]))
}
dsp.equal <- function(bgn, end) {
print(sav_txt_lst[["Headline"]][bgn:end])
print(glb_txt_lst[["Headline"]][bgn:end])
}
#sav_txt_lst <- glb_txt_lst; all.equal(sav_txt_lst, glb_txt_lst)
#all.equal(sav_txt_lst[["Headline"]][1:4200], glb_txt_lst[["Headline"]][1:4200])
#all.equal(sav_txt_lst[["Headline"]][1:2000], glb_txt_lst[["Headline"]][1:2000])
#all.equal(sav_txt_lst[["Headline"]][1:1000], glb_txt_lst[["Headline"]][1:1000])
#all.equal(sav_txt_lst[["Headline"]][1:500], glb_txt_lst[["Headline"]][1:500])
#all.equal(sav_txt_lst[["Headline"]][1:200], glb_txt_lst[["Headline"]][1:200])
#all.equal(sav_txt_lst[["Headline"]][1:100], glb_txt_lst[["Headline"]][1:100])
#chk.equal( 1, 100)
#chk.equal(51, 100)
#chk.equal(81, 100)
#chk.equal(81, 90)
#chk.equal(81, 85)
#chk.equal(86, 90)
#chk.equal(96, 100)
#dsp.equal(86, 90)
glb_txt_map_df <- read.csv("mytxt_map.csv", comment.char="#", strip.white=TRUE)
glb_txt_lst <- list();
print(sprintf("Building glb_txt_lst..."))
glb_txt_lst <- foreach(txt_var=glb_txt_vars) %dopar% {
# for (txt_var in glb_txt_vars) {
txt_vctr <- glb_allobs_df[, txt_var]
# myapply_txtmap shd be created as a tm_map::content_transformer ?
#print(glb_txt_map_df)
#txt_var=glb_txt_vars[3]; txt_vctr <- glb_txt_lst[[txt_var]]
#print(rex_str <- glb_txt_map_df[glb_txt_map_df$rex_str == "\\bWall St\\.", "rex_str"])
#print(rex_str <- glb_txt_map_df[grepl("du Pont", glb_txt_map_df$rex_str), "rex_str"])
#print(rex_str <- glb_txt_map_df[glb_txt_map_df$rpl_str == "versus", "rex_str"])
#print(tmp_vctr <- grep(rex_str, txt_vctr, value=TRUE, ignore.case=FALSE))
#ret_lst <- regexec(rex_str, txt_vctr, ignore.case=FALSE); ret_lst <- regmatches(txt_vctr, ret_lst); ret_vctr <- sapply(1:length(ret_lst), function(pos_ix) ifelse(length(ret_lst[[pos_ix]]) > 0, ret_lst[[pos_ix]], "")); print(ret_vctr <- ret_vctr[ret_vctr != ""])
#gsub(rex_str, glb_txt_map_df[glb_txt_map_df$rex_str == rex_str, "rpl_str"], tmp_vctr, ignore.case=FALSE)
#grep("Hong Hong", txt_vctr, value=TRUE)
txt_vctr <- myapply_txtmap(txt_vctr, ignore.case=FALSE)
}
names(glb_txt_lst) <- glb_txt_vars
for (txt_var in glb_txt_vars) {
print(sprintf("Remaining Acronyms in %s:", txt_var))
txt_vctr <- glb_txt_lst[[txt_var]]
print(tmp_vctr <- grep("[[:upper:]]\\.", txt_vctr, value=TRUE, ignore.case=FALSE))
}
for (txt_var in glb_txt_vars) {
re_str <- "\\b(Fort|Ft\\.|Hong|Las|Los|New|Puerto|Saint|San|St\\.)( |-)(\\w)+"
print(sprintf("Remaining #%s# terms in %s: ", re_str, txt_var))
txt_vctr <- glb_txt_lst[[txt_var]]
print(orderBy(~ -.n +pattern, subset(chk_pattern_freq(re_str, ignore.case=FALSE),
grepl("( |-)[[:upper:]]", pattern))))
print(" consider cleaning if relevant to problem domain; geography name; .n > 1")
#grep("New G", txt_vctr, value=TRUE, ignore.case=FALSE)
#grep("St\\. Wins", txt_vctr, value=TRUE, ignore.case=FALSE)
}
for (txt_var in glb_txt_vars) {
re_str <- "\\b(N|S|E|W|C)( |\\.)(\\w)+"
print(sprintf("Remaining #%s# terms in %s: ", re_str, txt_var))
txt_vctr <- glb_txt_lst[[txt_var]]
print(orderBy(~ -.n +pattern, subset(chk_pattern_freq(re_str, ignore.case=FALSE),
grepl(".", pattern))))
#grep("N Weaver", txt_vctr, value=TRUE, ignore.case=FALSE)
}
for (txt_var in glb_txt_vars) {
re_str <- "\\b(North|South|East|West|Central)( |\\.)(\\w)+"
print(sprintf("Remaining #%s# terms in %s: ", re_str, txt_var))
txt_vctr <- glb_txt_lst[[txt_var]]
print(orderBy(~ -.n +pattern, subset(chk_pattern_freq(re_str, ignore.case=FALSE),
grepl(".", pattern))))
#grep("Central (African|Bankers|Cast|Italy|Role|Spring)", txt_vctr, value=TRUE, ignore.case=FALSE)
#grep("East (Africa|Berlin|London|Poland|Rivals|Spring)", txt_vctr, value=TRUE, ignore.case=FALSE)
#grep("North (American|Korean|West)", txt_vctr, value=TRUE, ignore.case=FALSE)
#grep("South (Pacific|Street)", txt_vctr, value=TRUE, ignore.case=FALSE)
#grep("St\\. Martins", txt_vctr, value=TRUE, ignore.case=FALSE)
}
find_cmpnd_wrds <- function(txt_vctr) {
txt_corpus <- Corpus(VectorSource(txt_vctr))
txt_corpus <- tm_map(txt_corpus, tolower)
txt_corpus <- tm_map(txt_corpus, PlainTextDocument)
txt_corpus <- tm_map(txt_corpus, removePunctuation,
preserve_intra_word_dashes=TRUE)
full_Tf_DTM <- DocumentTermMatrix(txt_corpus,
control=list(weighting=weightTf))
print(" Full TermMatrix:"); print(full_Tf_DTM)
full_Tf_mtrx <- as.matrix(full_Tf_DTM)
rownames(full_Tf_mtrx) <- rownames(glb_allobs_df) # print undreadable otherwise
full_Tf_vctr <- colSums(full_Tf_mtrx)
names(full_Tf_vctr) <- dimnames(full_Tf_DTM)[[2]]
#grep("year", names(full_Tf_vctr), value=TRUE)
#which.max(full_Tf_mtrx[, "yearlong"])
full_Tf_df <- as.data.frame(full_Tf_vctr)
names(full_Tf_df) <- "Tf.full"
full_Tf_df$term <- rownames(full_Tf_df)
#full_Tf_df$freq.full <- colSums(full_Tf_mtrx != 0)
full_Tf_df <- orderBy(~ -Tf.full, full_Tf_df)
cmpnd_Tf_df <- full_Tf_df[grep("-", full_Tf_df$term, value=TRUE) ,]
filter_df <- read.csv("mytxt_compound.csv", comment.char="#", strip.white=TRUE)
cmpnd_Tf_df$filter <- FALSE
for (row_ix in 1:nrow(filter_df))
cmpnd_Tf_df[!cmpnd_Tf_df$filter, "filter"] <-
grepl(filter_df[row_ix, "rex_str"],
cmpnd_Tf_df[!cmpnd_Tf_df$filter, "term"], ignore.case=TRUE)
cmpnd_Tf_df <- subset(cmpnd_Tf_df, !filter)
# Bug in tm_map(txt_corpus, removePunctuation, preserve_intra_word_dashes=TRUE) ???
# "net-a-porter" gets converted to "net-aporter"
#grep("net-a-porter", txt_vctr, ignore.case=TRUE, value=TRUE)
#grep("maser-laser", txt_vctr, ignore.case=TRUE, value=TRUE)
#txt_corpus[[which(grepl("net-a-porter", txt_vctr, ignore.case=TRUE))]]
#grep("\\b(across|longer)-(\\w)", cmpnd_Tf_df$term, ignore.case=TRUE, value=TRUE)
#grep("(\\w)-(affected|term)\\b", cmpnd_Tf_df$term, ignore.case=TRUE, value=TRUE)
print(sprintf("nrow(cmpnd_Tf_df): %d", nrow(cmpnd_Tf_df)))
myprint_df(cmpnd_Tf_df)
}
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "process.text_reporting_compound_terms"), major.inc=FALSE)
for (txt_var in glb_txt_vars) {
print(sprintf("Remaining compound terms in %s: ", txt_var))
txt_vctr <- glb_txt_lst[[txt_var]]
# find_cmpnd_wrds(txt_vctr)
#grep("thirty-five", txt_vctr, ignore.case=TRUE, value=TRUE)
#rex_str <- glb_txt_map_df[grepl("hirty", glb_txt_map_df$rex_str), "rex_str"]
}
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "build.corpus"), major.inc=TRUE)
glb_corpus_lst <- list()
print(sprintf("Building glb_corpus_lst..."))
glb_corpus_lst <- foreach(txt_var=glb_txt_vars) %dopar% {
# for (txt_var in glb_txt_vars) {
txt_corpus <- Corpus(VectorSource(glb_txt_lst[[txt_var]]))
txt_corpus <- tm_map(txt_corpus, tolower) #nuppr
txt_corpus <- tm_map(txt_corpus, PlainTextDocument)
txt_corpus <- tm_map(txt_corpus, removePunctuation) #npnct<chr_ix>
# txt-corpus <- tm_map(txt_corpus, content_transformer(function(x, pattern) gsub(pattern, "", x))
# Not to be run in production
inspect_terms <- function() {
full_Tf_DTM <- DocumentTermMatrix(txt_corpus,
control=list(weighting=weightTf))
print(" Full TermMatrix:"); print(full_Tf_DTM)
full_Tf_mtrx <- as.matrix(full_Tf_DTM)
rownames(full_Tf_mtrx) <- rownames(glb_allobs_df) # print undreadable otherwise
full_Tf_vctr <- colSums(full_Tf_mtrx)
names(full_Tf_vctr) <- dimnames(full_Tf_DTM)[[2]]
#grep("year", names(full_Tf_vctr), value=TRUE)
#which.max(full_Tf_mtrx[, "yearlong"])
full_Tf_df <- as.data.frame(full_Tf_vctr)
names(full_Tf_df) <- "Tf.full"
full_Tf_df$term <- rownames(full_Tf_df)
#full_Tf_df$freq.full <- colSums(full_Tf_mtrx != 0)
full_Tf_df <- orderBy(~ -Tf.full +term, full_Tf_df)
print(myplot_histogram(full_Tf_df, "Tf.full"))
myprint_df(full_Tf_df)
#txt_corpus[[which(grepl("zun", txt_vctr, ignore.case=TRUE))]]
digit_terms_df <- subset(full_Tf_df, grepl("[[:digit:]]", term))
myprint_df(digit_terms_df)
return(full_Tf_df)
}
#print("RemovePunct:"); remove_punct_Tf_df <- inspect_terms()
txt_corpus <- tm_map(txt_corpus, removeWords,
c(glb_append_stop_words[[txt_var]],
stopwords("english"))) #nstopwrds
#print("StoppedWords:"); stopped_words_Tf_df <- inspect_terms()
txt_corpus <- tm_map(txt_corpus, stemDocument) #Features for lost information: Difference/ratio in density of full_TfIdf_DTM ???
#txt_corpus <- tm_map(txt_corpus, content_transformer(stemDocument))
#print("StemmedWords:"); stemmed_words_Tf_df <- inspect_terms()
#stemmed_stopped_Tf_df <- merge(stemmed_words_Tf_df, stopped_words_Tf_df, by="term", all=TRUE, suffixes=c(".stem", ".stop"))
#myprint_df(stemmed_stopped_Tf_df)
#print(subset(stemmed_stopped_Tf_df, grepl("compan", term)))
#glb_corpus_lst[[txt_var]] <- txt_corpus
}
names(glb_corpus_lst) <- glb_txt_vars
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "extract.DTM"), major.inc=TRUE)
glb_full_DTM_lst <- list(); glb_sprs_DTM_lst <- list();
for (txt_var in glb_txt_vars) {
print(sprintf("Extracting TfIDf terms for %s...", txt_var))
txt_corpus <- glb_corpus_lst[[txt_var]]
# full_Tf_DTM <- DocumentTermMatrix(txt_corpus,
# control=list(weighting=weightTf))
full_TfIdf_DTM <- DocumentTermMatrix(txt_corpus,
control=list(weighting=weightTfIdf))
sprs_TfIdf_DTM <- removeSparseTerms(full_TfIdf_DTM,
glb_sprs_thresholds[txt_var])
# glb_full_DTM_lst[[txt_var]] <- full_Tf_DTM
# glb_sprs_DTM_lst[[txt_var]] <- sprs_Tf_DTM
glb_full_DTM_lst[[txt_var]] <- full_TfIdf_DTM
glb_sprs_DTM_lst[[txt_var]] <- sprs_TfIdf_DTM
}
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "report.DTM"), major.inc=TRUE)
for (txt_var in glb_txt_vars) {
print(sprintf("Reporting TfIDf terms for %s...", txt_var))
full_TfIdf_DTM <- glb_full_DTM_lst[[txt_var]]
sprs_TfIdf_DTM <- glb_sprs_DTM_lst[[txt_var]]
print(" Full TermMatrix:"); print(full_TfIdf_DTM)
full_TfIdf_mtrx <- as.matrix(full_TfIdf_DTM)
rownames(full_TfIdf_mtrx) <- rownames(glb_allobs_df) # print undreadable otherwise
full_TfIdf_vctr <- colSums(full_TfIdf_mtrx)
names(full_TfIdf_vctr) <- dimnames(full_TfIdf_DTM)[[2]]
#grep("scene", names(full_TfIdf_vctr), value=TRUE)
#which.max(full_TfIdf_mtrx[, "yearlong"])
full_TfIdf_df <- as.data.frame(full_TfIdf_vctr)
names(full_TfIdf_df) <- "TfIdf.full"
full_TfIdf_df$term <- rownames(full_TfIdf_df)
full_TfIdf_df$freq.full <- colSums(full_TfIdf_mtrx != 0)
full_TfIdf_df <- orderBy(~ -TfIdf.full, full_TfIdf_df)
print(" Sparse TermMatrix:"); print(sprs_TfIdf_DTM)
sprs_TfIdf_vctr <- colSums(as.matrix(sprs_TfIdf_DTM))
names(sprs_TfIdf_vctr) <- dimnames(sprs_TfIdf_DTM)[[2]]
sprs_TfIdf_df <- as.data.frame(sprs_TfIdf_vctr)
names(sprs_TfIdf_df) <- "TfIdf.sprs"
sprs_TfIdf_df$term <- rownames(sprs_TfIdf_df)
sprs_TfIdf_df$freq.sprs <- colSums(as.matrix(sprs_TfIdf_DTM) != 0)
sprs_TfIdf_df <- orderBy(~ -TfIdf.sprs, sprs_TfIdf_df)
terms_TfIdf_df <- merge(full_TfIdf_df, sprs_TfIdf_df, all.x=TRUE)
terms_TfIdf_df$in.sprs <- !is.na(terms_TfIdf_df$freq.sprs)
plt_TfIdf_df <- subset(terms_TfIdf_df,
TfIdf.full >= min(terms_TfIdf_df$TfIdf.sprs, na.rm=TRUE))
plt_TfIdf_df$label <- ""
plt_TfIdf_df[is.na(plt_TfIdf_df$TfIdf.sprs), "label"] <-
plt_TfIdf_df[is.na(plt_TfIdf_df$TfIdf.sprs), "term"]
glb_important_terms[[txt_var]] <- union(glb_important_terms[[txt_var]],
plt_TfIdf_df[is.na(plt_TfIdf_df$TfIdf.sprs), "term"])
print(myplot_scatter(plt_TfIdf_df, "freq.full", "TfIdf.full",
colorcol_name="in.sprs") +
geom_text(aes(label=label), color="Black", size=3.5))
melt_TfIdf_df <- orderBy(~ -value, melt(terms_TfIdf_df, id.var="term"))
print(ggplot(melt_TfIdf_df, aes(value, color=variable)) + stat_ecdf() +
geom_hline(yintercept=glb_sprs_thresholds[txt_var],
linetype = "dotted"))
melt_TfIdf_df <- orderBy(~ -value,
melt(subset(terms_TfIdf_df, !is.na(TfIdf.sprs)), id.var="term"))
print(myplot_hbar(melt_TfIdf_df, "term", "value",
colorcol_name="variable"))
melt_TfIdf_df <- orderBy(~ -value,
melt(subset(terms_TfIdf_df, is.na(TfIdf.sprs)), id.var="term"))
print(myplot_hbar(head(melt_TfIdf_df, 10), "term", "value",
colorcol_name="variable"))
}
# sav_full_DTM_lst <- glb_full_DTM_lst
# sav_sprs_DTM_lst <- glb_sprs_DTM_lst
# print(identical(sav_glb_corpus_lst, glb_corpus_lst))
# print(all.equal(length(sav_glb_corpus_lst), length(glb_corpus_lst)))
# print(all.equal(names(sav_glb_corpus_lst), names(glb_corpus_lst)))
# print(all.equal(sav_glb_corpus_lst[["Headline"]], glb_corpus_lst[["Headline"]]))
# print(identical(sav_full_DTM_lst, glb_full_DTM_lst))
# print(identical(sav_sprs_DTM_lst, glb_sprs_DTM_lst))
rm(full_TfIdf_mtrx, full_TfIdf_df, melt_TfIdf_df, terms_TfIdf_df)
# Create txt features
if ((length(glb_txt_vars) > 1) &&
(length(unique(pfxs <- sapply(glb_txt_vars,
function(txt) toupper(substr(txt, 1, 1))))) < length(glb_txt_vars)))
stop("Prefixes for corpus freq terms not unique: ", pfxs)
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "bind.DTM"),
major.inc=TRUE)
for (txt_var in glb_txt_vars) {
print(sprintf("Binding DTM for %s...", txt_var))
txt_var_pfx <- toupper(substr(txt_var, 1, 1))
txt_X_df <- as.data.frame(as.matrix(glb_sprs_DTM_lst[[txt_var]]))
colnames(txt_X_df) <- paste(txt_var_pfx, ".T.",
make.names(colnames(txt_X_df)), sep="")
rownames(txt_X_df) <- rownames(glb_allobs_df) # warning otherwise
# plt_X_df <- cbind(txt_X_df, glb_allobs_df[, c(glb_id_var, glb_rsp_var)])
# print(myplot_box(df=plt_X_df, ycol_names="H.T.today", xcol_name=glb_rsp_var))
# log_X_df <- log(1 + txt_X_df)
# colnames(log_X_df) <- paste(colnames(txt_X_df), ".log", sep="")
# plt_X_df <- cbind(log_X_df, glb_allobs_df[, c(glb_id_var, glb_rsp_var)])
# print(myplot_box(df=plt_X_df, ycol_names="H.T.today.log", xcol_name=glb_rsp_var))
glb_allobs_df <- cbind(glb_allobs_df, txt_X_df) # TfIdf is normalized
#glb_allobs_df <- cbind(glb_allobs_df, log_X_df) # if using non-normalized metrics
}
#identical(chk_entity_df, glb_allobs_df)
#chk_entity_df <- glb_allobs_df
extract.features_chunk_df <- myadd_chunk(extract.features_chunk_df,
paste0("extract.features_", "bind.DXM"),
major.inc=TRUE)
#sav_allobs_df <- glb_allobs_df
glb_punct_vctr <- c("!", "\"", "#", "\\$", "%", "&", "'",
"\\(|\\)",# "\\(", "\\)",
"\\*", "\\+", ",", "-", "\\.", "/", ":", ";",
"<|>", # "<",
"=",
# ">",
"\\?", "@", "\\[", "\\\\", "\\]", "^", "_", "`",
"\\{", "\\|", "\\}", "~")
txt_X_df <- glb_allobs_df[, c(glb_id_var, ".rnorm"), FALSE]
txt_X_df <- foreach(txt_var=glb_txt_vars, .combine=cbind) %dopar% {
#for (txt_var in glb_txt_vars) {
print(sprintf("Binding DXM for %s...", txt_var))
txt_var_pfx <- toupper(substr(txt_var, 1, 1))
#txt_X_df <- glb_allobs_df[, c(glb_id_var, ".rnorm"), FALSE]
txt_full_DTM_mtrx <- as.matrix(glb_full_DTM_lst[[txt_var]])
rownames(txt_full_DTM_mtrx) <- rownames(glb_allobs_df) # print undreadable otherwise
#print(txt_full_DTM_mtrx[txt_full_DTM_mtrx[, "ebola"] != 0, "ebola"])
# Create <txt_var>.T.<term> for glb_important_terms
for (term in glb_important_terms[[txt_var]])
txt_X_df[, paste0(txt_var_pfx, ".T.", make.names(term))] <-
txt_full_DTM_mtrx[, term]
# Create <txt_var>.nwrds.log & .nwrds.unq.log
txt_X_df[, paste0(txt_var_pfx, ".nwrds.log")] <-
log(1 + mycount_pattern_occ("\\w+", glb_txt_lst[[txt_var]]))
txt_X_df[, paste0(txt_var_pfx, ".nwrds.unq.log")] <-
log(1 + rowSums(txt_full_DTM_mtrx != 0))
txt_X_df[, paste0(txt_var_pfx, ".sum.TfIdf")] <-
rowSums(txt_full_DTM_mtrx)
txt_X_df[, paste0(txt_var_pfx, ".ratio.sum.TfIdf.nwrds")] <-
txt_X_df[, paste0(txt_var_pfx, ".sum.TfIdf")] /
(exp(txt_X_df[, paste0(txt_var_pfx, ".nwrds.log")]) - 1)
txt_X_df[is.nan(txt_X_df[, paste0(txt_var_pfx, ".ratio.sum.TfIdf.nwrds")]),
paste0(txt_var_pfx, ".ratio.sum.TfIdf.nwrds")] <- 0
# Create <txt_var>.nchrs.log
txt_X_df[, paste0(txt_var_pfx, ".nchrs.log")] <-
log(1 + mycount_pattern_occ(".", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".nuppr.log")] <-
log(1 + mycount_pattern_occ("[[:upper:]]", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".ndgts.log")] <-
log(1 + mycount_pattern_occ("[[:digit:]]", glb_allobs_df[, txt_var]))
# Create <txt_var>.npnct?.log
# would this be faster if it's iterated over each row instead of
# each created column ???
for (punct_ix in 1:length(glb_punct_vctr)) {
# smp0 <- " "
# smp1 <- "! \" # $ % & ' ( ) * + , - . / : ; < = > ? @ [ \ ] ^ _ ` { | } ~"
# smp2 <- paste(smp1, smp1, sep=" ")
# print(sprintf("Testing %s pattern:", glb_punct_vctr[punct_ix]))
# results <- mycount_pattern_occ(glb_punct_vctr[punct_ix], c(smp0, smp1, smp2))
# names(results) <- NULL; print(results)
txt_X_df[,
paste0(txt_var_pfx, ".npnct", sprintf("%02d", punct_ix), ".log")] <-
log(1 + mycount_pattern_occ(glb_punct_vctr[punct_ix],
glb_allobs_df[, txt_var]))
}
# print(head(glb_allobs_df[glb_allobs_df[, "A.npnct23.log"] > 0,
# c("UniqueID", "Popular", "Abstract", "A.npnct23.log")]))
# Create <txt_var>.nstopwrds.log & <txt_var>ratio.nstopwrds.nwrds
stop_words_rex_str <- paste0("\\b(", paste0(c(glb_append_stop_words[[txt_var]],
stopwords("english")), collapse="|"),
")\\b")
txt_X_df[, paste0(txt_var_pfx, ".nstopwrds", ".log")] <-
log(1 + mycount_pattern_occ(stop_words_rex_str, glb_txt_lst[[txt_var]]))
txt_X_df[, paste0(txt_var_pfx, ".ratio.nstopwrds.nwrds")] <-
exp(txt_X_df[, paste0(txt_var_pfx, ".nstopwrds", ".log")] -
txt_X_df[, paste0(txt_var_pfx, ".nwrds", ".log")])
# Create <txt_var>.P.http
txt_X_df[, paste(txt_var_pfx, ".P.http", sep="")] <-
as.integer(0 + mycount_pattern_occ("http", glb_allobs_df[, txt_var]))
# Create user-specified pattern vectors
# <txt_var>.P.year.colon
txt_X_df[, paste0(txt_var_pfx, ".P.year.colon")] <-
as.integer(0 + mycount_pattern_occ("[0-9]{4}:", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.daily.clip.report")] <-
as.integer(0 + mycount_pattern_occ("Daily Clip Report", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.fashion.week")] <-
as.integer(0 + mycount_pattern_occ("Fashion Week", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.first.draft")] <-
as.integer(0 + mycount_pattern_occ("First Draft", glb_allobs_df[, txt_var]))
#sum(mycount_pattern_occ("Metropolitan Diary:", glb_allobs_df$Abstract) > 0)
if (txt_var %in% c("Snippet", "Abstract")) {
txt_X_df[, paste0(txt_var_pfx, ".P.metropolitan.diary.colon")] <-
as.integer(0 + mycount_pattern_occ("Metropolitan Diary:",
glb_allobs_df[, txt_var]))
}
#sum(mycount_pattern_occ("[0-9]{4}:", glb_allobs_df$Headline) > 0)
#sum(mycount_pattern_occ("Quandary(.*)(?=:)", glb_allobs_df$Headline, perl=TRUE) > 0)
#sum(mycount_pattern_occ("No Comment(.*):", glb_allobs_df$Headline) > 0)
#sum(mycount_pattern_occ("Friday Night Music:", glb_allobs_df$Headline) > 0)
if (txt_var %in% c("Headline")) {
txt_X_df[, paste0(txt_var_pfx, ".P.facts.figures")] <-
as.integer(0 + mycount_pattern_occ("Facts & Figures:", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.friday.night.music")] <-
as.integer(0 + mycount_pattern_occ("Friday Night Music", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.no.comment.colon")] <-
as.integer(0 + mycount_pattern_occ("No Comment(.*):", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.on.this.day")] <-
as.integer(0 + mycount_pattern_occ("On This Day", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.quandary")] <-
as.integer(0 + mycount_pattern_occ("Quandary(.*)(?=:)", glb_allobs_df[, txt_var], perl=TRUE))
txt_X_df[, paste0(txt_var_pfx, ".P.readers.respond")] <-
as.integer(0 + mycount_pattern_occ("Readers Respond", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.recap.colon")] <-
as.integer(0 + mycount_pattern_occ("Recap:", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.s.notebook")] <-
as.integer(0 + mycount_pattern_occ("s Notebook", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.today.in.politic")] <-
as.integer(0 + mycount_pattern_occ("Today in Politic", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.today.in.smallbusiness")] <-
as.integer(0 + mycount_pattern_occ("Today in Small Business:", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.verbatim.colon")] <-
as.integer(0 + mycount_pattern_occ("Verbatim:", glb_allobs_df[, txt_var]))
txt_X_df[, paste0(txt_var_pfx, ".P.what.we.are")] <-
as.integer(0 + mycount_pattern_occ("What We're", glb_allobs_df[, txt_var]))
}
#summary(glb_allobs_df[ ,grep("P.on.this.day", names(glb_allobs_df), value=TRUE)])
txt_X_df <- subset(txt_X_df, select=-.rnorm)
txt_X_df <- txt_X_df[, -grep(glb_id_var, names(txt_X_df), fixed=TRUE), FALSE]
#glb_allobs_df <- cbind(glb_allobs_df, txt_X_df)
}
glb_allobs_df <- cbind(glb_allobs_df, txt_X_df)
#myplot_box(glb_allobs_df, "A.sum.TfIdf", glb_rsp_var)
# Generate summaries
# print(summary(glb_allobs_df))
# print(sapply(names(glb_allobs_df), function(col) sum(is.na(glb_allobs_df[, col]))))
# print(summary(glb_trnobs_df))
# print(sapply(names(glb_trnobs_df), function(col) sum(is.na(glb_trnobs_df[, col]))))
# print(summary(glb_newobs_df))
# print(sapply(names(glb_newobs_df), function(col) sum(is.na(glb_newobs_df[, col]))))
glb_exclude_vars_as_features <- union(glb_exclude_vars_as_features,
glb_txt_vars)
rm(log_X_df, txt_X_df)
}
# print(sapply(names(glb_trnobs_df), function(col) sum(is.na(glb_trnobs_df[, col]))))
# print(sapply(names(glb_newobs_df), function(col) sum(is.na(glb_newobs_df[, col]))))
# print(myplot_scatter(glb_trnobs_df, "<col1_name>", "<col2_name>", smooth=TRUE))
rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
glb_full_DTM_lst, glb_sprs_DTM_lst, txt_corpus, txt_vctr)## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'corpus_lst' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'full_TfIdf_DTM' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'full_TfIdf_vctr' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'glb_full_DTM_lst' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'glb_sprs_DTM_lst' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'txt_corpus' not found## Warning in rm(corpus_lst, full_TfIdf_DTM, full_TfIdf_vctr,
## glb_full_DTM_lst, : object 'txt_vctr' not foundextract.features_chunk_df <- myadd_chunk(extract.features_chunk_df, "extract.features_end",
major.inc=TRUE)## label step_major step_minor bgn end
## 2 extract.features_factorize.str.vars 2 0 64.155 64.338
## 3 extract.features_end 3 0 64.339 NA
## elapsed
## 2 0.183
## 3 NAmyplt_chunk(extract.features_chunk_df)## label step_major step_minor bgn end
## 2 extract.features_factorize.str.vars 2 0 64.155 64.338
## 1 extract.features_bgn 1 0 64.144 64.154
## elapsed duration
## 2 0.183 0.183
## 1 0.010 0.010
## [1] "Total Elapsed Time: 64.338 secs"
# if (glb_save_envir)
# save(glb_feats_df,
# glb_allobs_df, #glb_trnobs_df, glb_fitobs_df, glb_OOBobs_df, glb_newobs_df,
# file=paste0(glb_out_pfx, "extract_features_dsk.RData"))
# load(paste0(glb_out_pfx, "extract_features_dsk.RData"))
replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"data.training.all","data.new")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
glb_chunks_df <- myadd_chunk(glb_chunks_df, "cluster.data", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 6 extract.features 3 0 64.137 65.84 1.703
## 7 cluster.data 4 0 65.840 NA NAStep 4.0: cluster data
if (glb_cluster) {
require(proxy)
#require(hash)
require(dynamicTreeCut)
# glb_hash <- hash(key=unique(glb_allobs_df$myCategory),
# values=1:length(unique(glb_allobs_df$myCategory)))
# glb_hash_lst <- hash(key=unique(glb_allobs_df$myCategory),
# values=1:length(unique(glb_allobs_df$myCategory)))
#stophere; sav_allobs_df <- glb_allobs_df;
print("Clustering features: ")
print(cluster_vars <- grep("[HSA]\\.[PT]\\.", names(glb_allobs_df), value=TRUE))
#print(cluster_vars <- grep("[HSA]\\.", names(glb_allobs_df), value=TRUE))
glb_allobs_df$.clusterid <- 1
#print(max(table(glb_allobs_df$myCategory.fctr) / 20))
for (myCategory in c("##", "Business#Business Day#Dealbook", "OpEd#Opinion#",
"Styles#U.S.#", "Business#Technology#", "Science#Health#",
"Culture#Arts#")) {
ctgry_allobs_df <- glb_allobs_df[glb_allobs_df$myCategory == myCategory, ]
dstns_dist <- dist(ctgry_allobs_df[, cluster_vars], method = "cosine")
dstns_mtrx <- as.matrix(dstns_dist)
print(sprintf("max distance(%0.4f) pair:", max(dstns_mtrx)))
row_ix <- ceiling(which.max(dstns_mtrx) / ncol(dstns_mtrx))
col_ix <- which.max(dstns_mtrx[row_ix, ])
print(ctgry_allobs_df[c(row_ix, col_ix),
c("UniqueID", "Popular", "myCategory", "Headline", cluster_vars)])
min_dstns_mtrx <- dstns_mtrx
diag(min_dstns_mtrx) <- 1
print(sprintf("min distance(%0.4f) pair:", min(min_dstns_mtrx)))
row_ix <- ceiling(which.min(min_dstns_mtrx) / ncol(min_dstns_mtrx))
col_ix <- which.min(min_dstns_mtrx[row_ix, ])
print(ctgry_allobs_df[c(row_ix, col_ix),
c("UniqueID", "Popular", "myCategory", "Headline", cluster_vars)])
clusters <- hclust(dstns_dist, method = "ward.D2")
#plot(clusters, labels=NULL, hang=-1)
myplclust(clusters, lab.col=unclass(ctgry_allobs_df[, glb_rsp_var]))
#clusterGroups = cutree(clusters, k=7)
clusterGroups <- cutreeDynamic(clusters, minClusterSize=20, method="tree", deepSplit=0)
# Unassigned groups are labeled 0; the largest group has label 1
table(clusterGroups, ctgry_allobs_df[, glb_rsp_var], useNA="ifany")
#print(ctgry_allobs_df[which(clusterGroups == 1), c("UniqueID", "Popular", "Headline")])
#print(ctgry_allobs_df[(clusterGroups == 1) & !is.na(ctgry_allobs_df$Popular) & (ctgry_allobs_df$Popular==1), c("UniqueID", "Popular", "Headline")])
clusterGroups[clusterGroups == 0] <- 1
table(clusterGroups, ctgry_allobs_df[, glb_rsp_var], useNA="ifany")
#summary(factor(clusterGroups))
# clusterGroups <- clusterGroups +
# 100 * # has to be > max(table(glb_allobs_df$myCategory.fctr) / minClusterSize=20)
# which(levels(glb_allobs_df$myCategory.fctr) == myCategory)
# table(clusterGroups, ctgry_allobs_df[, glb_rsp_var], useNA="ifany")
# add to glb_allobs_df - then split the data again
glb_allobs_df[glb_allobs_df$myCategory==myCategory,]$.clusterid <- clusterGroups
#print(unique(glb_allobs_df$.clusterid))
#print(glb_feats_df[glb_feats_df$id == ".clusterid.fctr", ])
}
ctgry_xtab_df <- orderBy(reformulate(c("-", ".n")),
mycreate_sqlxtab_df(glb_allobs_df,
c("myCategory", ".clusterid", glb_rsp_var)))
ctgry_cast_df <- orderBy(~ -Y -NA, dcast(ctgry_xtab_df,
myCategory + .clusterid ~
Popular.fctr, sum, value.var=".n"))
print(ctgry_cast_df)
#print(orderBy(~ myCategory -Y -NA, ctgry_cast_df))
# write.table(ctgry_cast_df, paste0(glb_out_pfx, "ctgry_clst.csv"),
# row.names=FALSE)
print(ctgry_sum_tbl <- table(glb_allobs_df$myCategory, glb_allobs_df$.clusterid,
glb_allobs_df[, glb_rsp_var],
useNA="ifany"))
# dsp_obs(.clusterid=1, myCategory="OpEd#Opinion#",
# cols=c("UniqueID", "Popular", "myCategory", ".clusterid", "Headline"),
# all=TRUE)
glb_allobs_df$.clusterid.fctr <- as.factor(glb_allobs_df$.clusterid)
glb_exclude_vars_as_features <- c(glb_exclude_vars_as_features,
".clusterid")
glb_interaction_only_features["myCategory.fctr"] <- c(".clusterid.fctr")
glb_exclude_vars_as_features <- c(glb_exclude_vars_as_features,
cluster_vars)
}
# Re-partition
glb_trnobs_df <- subset(glb_allobs_df, .src == "Train")
glb_newobs_df <- subset(glb_allobs_df, .src == "Test")
glb_chunks_df <- myadd_chunk(glb_chunks_df, "select.features", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 7 cluster.data 4 0 65.840 66.701 0.861
## 8 select.features 5 0 66.701 NA NAStep 5.0: select features
print(glb_feats_df <- myselect_features(entity_df=glb_trnobs_df,
exclude_vars_as_features=glb_exclude_vars_as_features,
rsp_var=glb_rsp_var))## id cor.y exclude.as.feat
## Industry.my.fctr Industry.my.fctr -0.376894239 0
## Age Age 0.279553016 0
## Education.my.fctr Education.my.fctr -0.095424564 0
## PeopleInHousehold PeopleInHousehold -0.086274719 0
## Sex.fctr Sex.fctr 0.061614520 0
## Married.my.fctr Married.my.fctr 0.031303760 0
## Country.my.fctr Country.my.fctr -0.024685468 1
## CountryOfBirthCode CountryOfBirthCode -0.023914164 1
## Hispanic Hispanic -0.020230546 0
## Citizenship.fctr Citizenship.fctr -0.019730560 0
## State.fctr State.fctr -0.017050319 1
## Region.fctr Region.fctr 0.016534173 0
## MetroArea.my.fctr MetroArea.my.fctr -0.004844306 1
## MetroAreaCode MetroAreaCode -0.004243514 1
## Race.fctr Race.fctr -0.002325650 0
## .rnorm .rnorm 0.002199196 0
## cor.y.abs
## Industry.my.fctr 0.376894239
## Age 0.279553016
## Education.my.fctr 0.095424564
## PeopleInHousehold 0.086274719
## Sex.fctr 0.061614520
## Married.my.fctr 0.031303760
## Country.my.fctr 0.024685468
## CountryOfBirthCode 0.023914164
## Hispanic 0.020230546
## Citizenship.fctr 0.019730560
## State.fctr 0.017050319
## Region.fctr 0.016534173
## MetroArea.my.fctr 0.004844306
## MetroAreaCode 0.004243514
## Race.fctr 0.002325650
## .rnorm 0.002199196# sav_feats_df <- glb_feats_df; glb_feats_df <- sav_feats_df
print(glb_feats_df <- orderBy(~-cor.y,
myfind_cor_features(feats_df=glb_feats_df, obs_df=glb_trnobs_df,
rsp_var=glb_rsp_var)))## id cor.y exclude.as.feat cor.y.abs cor.high.X
## 2 Age 0.279553016 0 0.279553016 NA
## 15 Sex.fctr 0.061614520 0 0.061614520 NA
## 9 Married.my.fctr 0.031303760 0 0.031303760 NA
## 14 Region.fctr 0.016534173 0 0.016534173 NA
## 1 .rnorm 0.002199196 0 0.002199196 NA
## 13 Race.fctr -0.002325650 0 0.002325650 NA
## 11 MetroAreaCode -0.004243514 1 0.004243514 NA
## 10 MetroArea.my.fctr -0.004844306 1 0.004844306 NA
## 16 State.fctr -0.017050319 1 0.017050319 NA
## 3 Citizenship.fctr -0.019730560 0 0.019730560 NA
## 7 Hispanic -0.020230546 0 0.020230546 NA
## 5 CountryOfBirthCode -0.023914164 1 0.023914164 NA
## 4 Country.my.fctr -0.024685468 1 0.024685468 NA
## 12 PeopleInHousehold -0.086274719 0 0.086274719 NA
## 6 Education.my.fctr -0.095424564 0 0.095424564 NA
## 8 Industry.my.fctr -0.376894239 0 0.376894239 NA
## freqRatio percentUnique zeroVar nzv myNearZV is.cor.y.abs.low
## 2 1.089125 0.063499284 FALSE FALSE FALSE FALSE
## 15 1.097632 0.001895501 FALSE FALSE FALSE FALSE
## 9 1.797322 0.004738753 FALSE FALSE FALSE FALSE
## 14 1.274445 0.003791002 FALSE FALSE FALSE FALSE
## 1 1.000000 99.687242330 FALSE FALSE FALSE FALSE
## 13 7.990088 0.005686503 FALSE FALSE FALSE FALSE
## 11 1.303351 0.250206136 FALSE FALSE FALSE FALSE
## 10 6.253270 0.251153886 FALSE FALSE FALSE FALSE
## 16 1.706890 0.048335276 FALSE FALSE FALSE FALSE
## 3 12.901057 0.002843252 FALSE FALSE FALSE FALSE
## 7 7.182474 0.001895501 FALSE FALSE FALSE FALSE
## 5 24.005593 0.152587833 FALSE TRUE FALSE FALSE
## 4 24.005593 0.137423825 FALSE TRUE FALSE FALSE
## 12 1.820761 0.014216258 FALSE FALSE FALSE FALSE
## 6 1.590536 0.007582004 FALSE FALSE FALSE FALSE
## 8 2.615103 0.014216258 FALSE FALSE FALSE FALSE#subset(glb_feats_df, id %in% c("A.nuppr.log", "S.nuppr.log"))
print(myplot_scatter(glb_feats_df, "percentUnique", "freqRatio",
colorcol_name="myNearZV", jitter=TRUE) +
geom_point(aes(shape=nzv)) + xlim(-5, 25))## Warning in myplot_scatter(glb_feats_df, "percentUnique", "freqRatio",
## colorcol_name = "myNearZV", : converting myNearZV to class:factor## Warning in loop_apply(n, do.ply): Removed 1 rows containing missing values
## (geom_point).## Warning in loop_apply(n, do.ply): Removed 1 rows containing missing values
## (geom_point).## Warning in loop_apply(n, do.ply): Removed 1 rows containing missing values
## (geom_point).
print(subset(glb_feats_df, myNearZV))## [1] id cor.y exclude.as.feat cor.y.abs
## [5] cor.high.X freqRatio percentUnique zeroVar
## [9] nzv myNearZV is.cor.y.abs.low
## <0 rows> (or 0-length row.names)glb_allobs_df <- glb_allobs_df[, setdiff(names(glb_allobs_df),
subset(glb_feats_df, myNearZV)$id)]
if (!is.null(glb_interaction_only_features))
glb_feats_df[glb_feats_df$id %in% glb_interaction_only_features, "interaction.feat"] <-
names(glb_interaction_only_features) else
glb_feats_df$interaction.feat <- NA
mycheck_problem_data(glb_allobs_df, terminate = TRUE)## [1] "numeric data missing in : "
## MetroAreaCode EmploymentStatus.fctr
## 34238 25789
## [1] "numeric data w/ 0s in : "
## Age Hispanic
## 1283 113008
## [1] "numeric data w/ Infs in : "
## named integer(0)
## [1] "numeric data w/ NaNs in : "
## named integer(0)
## [1] "string data missing in : "
## Region State Married Sex
## 0 0 NA 0
## Education Race Citizenship EmploymentStatus
## NA 0 0 NA
## Industry .rownames MetroArea Country
## NA 0 NA NA
## Married.my Education.my Industry.my MetroArea.my
## 0 0 0 0
## Country.my
## 0# glb_allobs_df %>% filter(is.na(Married.fctr)) %>% tbl_df()
# glb_allobs_df %>% count(Married.fctr)
# levels(glb_allobs_df$Married.fctr)
glb_chunks_df <- myadd_chunk(glb_chunks_df, "partition.data.training", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 8 select.features 5 0 66.701 74.324 7.624
## 9 partition.data.training 6 0 74.325 NA NAStep 6.0: partition data training
if (all(is.na(glb_newobs_df[, glb_rsp_var]))) {
require(caTools)
set.seed(glb_split_sample.seed)
split <- sample.split(glb_trnobs_df[, glb_rsp_var_raw],
SplitRatio=1 - (nrow(glb_newobs_df) * 1.1 / nrow(glb_trnobs_df)))
glb_fitobs_df <- glb_trnobs_df[split, ]
glb_OOBobs_df <- glb_trnobs_df[!split ,]
} else {
print(sprintf("Newdata contains non-NA data for %s; setting OOB to Newdata",
glb_rsp_var))
glb_fitobs_df <- glb_trnobs_df; glb_OOBobs_df <- glb_newobs_df
}## Loading required package: caToolsif (!is.null(glb_max_fitent_obs) && (nrow(glb_fitobs_df) > glb_max_fitent_obs)) {
warning("glb_fitobs_df restricted to glb_max_fitent_obs: ",
format(glb_max_fitent_obs, big.mark=","))
org_fitent_df <- glb_fitobs_df
glb_fitobs_df <-
org_fitent_df[split <- sample.split(org_fitent_df[, glb_rsp_var_raw],
SplitRatio=glb_max_fitent_obs), ]
org_fitent_df <- NULL
}
glb_allobs_df$.lcn <- ""
glb_allobs_df[glb_allobs_df[, glb_id_var] %in%
glb_fitobs_df[, glb_id_var], ".lcn"] <- "Fit"
glb_allobs_df[glb_allobs_df[, glb_id_var] %in%
glb_OOBobs_df[, glb_id_var], ".lcn"] <- "OOB"
dsp_class_dstrb <- function(obs_df, location_var, partition_var) {
xtab_df <- mycreate_xtab_df(obs_df, c(location_var, partition_var))
rownames(xtab_df) <- xtab_df[, location_var]
xtab_df <- xtab_df[, -grepl(location_var, names(xtab_df))]
print(xtab_df)
print(xtab_df / rowSums(xtab_df, na.rm=TRUE))
}
# Ensure proper splits by glb_rsp_var_raw & user-specified feature for OOB vs. new
if (!is.null(glb_category_vars)) {
if (glb_is_classification)
dsp_class_dstrb(glb_allobs_df, ".lcn", glb_rsp_var_raw)
newent_ctgry_df <- mycreate_sqlxtab_df(subset(glb_allobs_df, .src == "Test"),
glb_category_vars)
OOBobs_ctgry_df <- mycreate_sqlxtab_df(subset(glb_allobs_df, .lcn == "OOB"),
glb_category_vars)
glb_ctgry_df <- merge(newent_ctgry_df, OOBobs_ctgry_df, by=glb_category_vars
, all=TRUE, suffixes=c(".Tst", ".OOB"))
glb_ctgry_df$.freqRatio.Tst <- glb_ctgry_df$.n.Tst / sum(glb_ctgry_df$.n.Tst, na.rm=TRUE)
glb_ctgry_df$.freqRatio.OOB <- glb_ctgry_df$.n.OOB / sum(glb_ctgry_df$.n.OOB, na.rm=TRUE)
print(orderBy(~-.freqRatio.Tst-.freqRatio.OOB, glb_ctgry_df))
}
# Run this line by line
print("glb_feats_df:"); print(dim(glb_feats_df))## [1] "glb_feats_df:"## [1] 16 12sav_feats_df <- glb_feats_df
glb_feats_df <- sav_feats_df
glb_feats_df[, "rsp_var_raw"] <- FALSE
glb_feats_df[glb_feats_df$id == glb_rsp_var_raw, "rsp_var_raw"] <- TRUE
glb_feats_df$exclude.as.feat <- (glb_feats_df$exclude.as.feat == 1)
if (!is.null(glb_id_var) && glb_id_var != ".rownames")
glb_feats_df[glb_feats_df$id %in% glb_id_var, "id_var"] <- TRUE
add_feats_df <- data.frame(id=glb_rsp_var, exclude.as.feat=TRUE, rsp_var=TRUE)
row.names(add_feats_df) <- add_feats_df$id; print(add_feats_df)## id exclude.as.feat rsp_var
## EmploymentStatus.fctr EmploymentStatus.fctr TRUE TRUEglb_feats_df <- myrbind_df(glb_feats_df, add_feats_df)
if (glb_id_var != ".rownames")
print(subset(glb_feats_df, rsp_var_raw | rsp_var | id_var)) else
print(subset(glb_feats_df, rsp_var_raw | rsp_var)) ## id cor.y exclude.as.feat
## EmploymentStatus.fctr EmploymentStatus.fctr NA TRUE
## cor.y.abs cor.high.X freqRatio percentUnique zeroVar
## EmploymentStatus.fctr NA NA NA NA NA
## nzv myNearZV is.cor.y.abs.low interaction.feat
## EmploymentStatus.fctr NA NA NA NA
## rsp_var_raw rsp_var
## EmploymentStatus.fctr NA TRUEprint("glb_feats_df vs. glb_allobs_df: "); ## [1] "glb_feats_df vs. glb_allobs_df: "print(setdiff(glb_feats_df$id, names(glb_allobs_df)))## character(0)print("glb_allobs_df vs. glb_feats_df: "); ## [1] "glb_allobs_df vs. glb_feats_df: "# Ensure these are only chr vars
print(setdiff(setdiff(names(glb_allobs_df), glb_feats_df$id),
myfind_chr_cols_df(glb_allobs_df)))## character(0)#print(setdiff(setdiff(names(glb_allobs_df), glb_exclude_vars_as_features),
# glb_feats_df$id))
print("glb_allobs_df: "); print(dim(glb_allobs_df))## [1] "glb_allobs_df: "## [1] 131302 36print("glb_trnobs_df: "); print(dim(glb_trnobs_df))## [1] "glb_trnobs_df: "## [1] 105513 35print("glb_fitobs_df: "); print(dim(glb_fitobs_df))## [1] "glb_fitobs_df: "## [1] 77145 35print("glb_OOBobs_df: "); print(dim(glb_OOBobs_df))## [1] "glb_OOBobs_df: "## [1] 28368 35print("glb_newobs_df: "); print(dim(glb_newobs_df))## [1] "glb_newobs_df: "## [1] 25789 35# # Does not handle NULL or length(glb_id_var) > 1
# glb_allobs_df$.src.trn <- 0
# glb_allobs_df[glb_allobs_df[, glb_id_var] %in% glb_trnobs_df[, glb_id_var],
# ".src.trn"] <- 1
# glb_allobs_df$.src.fit <- 0
# glb_allobs_df[glb_allobs_df[, glb_id_var] %in% glb_fitobs_df[, glb_id_var],
# ".src.fit"] <- 1
# glb_allobs_df$.src.OOB <- 0
# glb_allobs_df[glb_allobs_df[, glb_id_var] %in% glb_OOBobs_df[, glb_id_var],
# ".src.OOB"] <- 1
# glb_allobs_df$.src.new <- 0
# glb_allobs_df[glb_allobs_df[, glb_id_var] %in% glb_newobs_df[, glb_id_var],
# ".src.new"] <- 1
# #print(unique(glb_allobs_df[, ".src.trn"]))
# write_cols <- c(glb_feats_df$id,
# ".src.trn", ".src.fit", ".src.OOB", ".src.new")
# glb_allobs_df <- glb_allobs_df[, write_cols]
#
# tmp_feats_df <- glb_feats_df
# tmp_entity_df <- glb_allobs_df
if (glb_save_envir)
save(glb_feats_df,
glb_allobs_df, #glb_trnobs_df, glb_fitobs_df, glb_OOBobs_df, glb_newobs_df,
file=paste0(glb_out_pfx, "blddfs_dsk.RData"))
# load(paste0(glb_out_pfx, "blddfs_dsk.RData"))
# if (!all.equal(tmp_feats_df, glb_feats_df))
# stop("glb_feats_df r/w not working")
# if (!all.equal(tmp_entity_df, glb_allobs_df))
# stop("glb_allobs_df r/w not working")
rm(split)
glb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.models", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 9 partition.data.training 6 0 74.325 75.172 0.847
## 10 fit.models 7 0 75.172 NA NAStep 7.0: fit models
# load(paste0(glb_out_pfx, "dsk.RData"))
# keep_cols <- setdiff(names(glb_allobs_df),
# grep("^.src", names(glb_allobs_df), value=TRUE))
# glb_trnobs_df <- glb_allobs_df[glb_allobs_df$.src.trn == 1, keep_cols]
# glb_fitobs_df <- glb_allobs_df[glb_allobs_df$.src.fit == 1, keep_cols]
# glb_OOBobs_df <- glb_allobs_df[glb_allobs_df$.src.OOB == 1, keep_cols]
# glb_newobs_df <- glb_allobs_df[glb_allobs_df$.src.new == 1, keep_cols]
#
# glb_models_lst <- list(); glb_models_df <- data.frame()
#
if (glb_is_classification && glb_is_binomial &&
(length(unique(glb_fitobs_df[, glb_rsp_var])) < 2))
stop("glb_fitobs_df$", glb_rsp_var, ": contains less than 2 unique values: ",
paste0(unique(glb_fitobs_df[, glb_rsp_var]), collapse=", "))
max_cor_y_x_vars <- orderBy(~ -cor.y.abs,
subset(glb_feats_df, (exclude.as.feat == 0) & !is.cor.y.abs.low &
is.na(cor.high.X)))[1:2, "id"]
# while(length(max_cor_y_x_vars) < 2) {
# max_cor_y_x_vars <- c(max_cor_y_x_vars, orderBy(~ -cor.y.abs,
# subset(glb_feats_df, (exclude.as.feat == 0) & !is.cor.y.abs.low))[3, "id"])
# }
if (!is.null(glb_Baseline_mdl_var)) {
if ((max_cor_y_x_vars[1] != glb_Baseline_mdl_var) &
(glb_feats_df[max_cor_y_x_vars[1], "cor.y.abs"] >
glb_feats_df[glb_Baseline_mdl_var, "cor.y.abs"]))
stop(max_cor_y_x_vars[1], " has a lower correlation with ", glb_rsp_var,
" than the Baseline var: ", glb_Baseline_mdl_var)
}
glb_model_type <- ifelse(glb_is_regression, "regression", "classification")
# Baseline
if (!is.null(glb_Baseline_mdl_var))
ret_lst <- myfit_mdl_fn(model_id="Baseline", model_method="mybaseln_classfr",
indep_vars_vctr=glb_Baseline_mdl_var,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df)
# Most Frequent Outcome "MFO" model: mean(y) for regression
# Not using caret's nullModel since model stats not avl
# Cannot use rpart for multinomial classification since it predicts non-MFO
ret_lst <- myfit_mdl(model_id="MFO",
model_method=ifelse(glb_is_regression, "lm", "myMFO_classfr"),
model_type=glb_model_type,
indep_vars_vctr=".rnorm",
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df)## [1] "fitting model: MFO.myMFO_classfr"
## [1] " indep_vars: .rnorm"
## Fitting parameter = none on full training set
## [1] "in MFO.Classifier$fit"
## [1] "unique.vals:"
## [1] Disabled Employed Not.in.Labor.Force
## [4] Retired Unemployed
## Levels: Disabled Employed Not.in.Labor.Force Retired Unemployed
## [1] "unique.prob:"
## y
## Employed Retired Not.in.Labor.Force
## 0.58508004 0.17645991 0.14449413
## Disabled Unemployed
## 0.05413183 0.03983408
## [1] "MFO.val:"
## [1] "Employed"
## Length Class Mode
## unique.vals 5 factor numeric
## unique.prob 5 -none- numeric
## MFO.val 1 -none- character
## x.names 1 -none- character
## xNames 1 -none- character
## problemType 1 -none- character
## tuneValue 1 data.frame list
## obsLevels 5 -none- character
## [1] " calling mypredict_mdl for fit:"
## [1] "entr MFO.Classifier$predict"
## [1] "exit MFO.Classifier$predict"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 4176 0 0
## Employed 0 45136 0 0
## Not.in.Labor.Force 0 11147 0 0
## Retired 0 13613 0 0
## Unemployed 0 3073 0 0
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.5850800 NA 0.5815936 0.5885601 0.5850800
## AccuracyPValue McnemarPValue
## 0.5015403 NaN
## [1] " calling mypredict_mdl for OOB:"
## [1] "entr MFO.Classifier$predict"
## [1] "exit MFO.Classifier$predict"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 1536 0 0
## Employed 0 16597 0 0
## Not.in.Labor.Force 0 4099 0 0
## Retired 0 5006 0 0
## Unemployed 0 1130 0 0
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.5850606 0.0000000 0.5793009 0.5908029 0.5850606
## AccuracyPValue McnemarPValue
## 0.5025399 NaN
## model_id model_method feats max.nTuningRuns
## 1 MFO.myMFO_classfr myMFO_classfr .rnorm 0
## min.elapsedtime.everything min.elapsedtime.final max.Accuracy.fit
## 1 0.471 0.012 0.58508
## max.AccuracyLower.fit max.AccuracyUpper.fit max.Kappa.fit
## 1 0.5815936 0.5885601 NA
## max.Accuracy.OOB max.AccuracyLower.OOB max.AccuracyUpper.OOB
## 1 0.5850606 0.5793009 0.5908029
## max.Kappa.OOB
## 1 0if (glb_is_classification)
# "random" model - only for classification;
# none needed for regression since it is same as MFO
ret_lst <- myfit_mdl(model_id="Random", model_method="myrandom_classfr",
model_type=glb_model_type,
indep_vars_vctr=".rnorm",
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df)## [1] "fitting model: Random.myrandom_classfr"
## [1] " indep_vars: .rnorm"
## Fitting parameter = none on full training set
## Length Class Mode
## unique.vals 5 factor numeric
## unique.prob 5 table numeric
## xNames 1 -none- character
## problemType 1 -none- character
## tuneValue 1 data.frame list
## obsLevels 5 -none- character
## [1] " calling mypredict_mdl for fit:"## Warning in sum(ni[1:m] * nj[1:m]): integer overflow - use
## sum(as.numeric(.))## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 214 2428 616 749
## Employed 2512 26320 6565 7896
## Not.in.Labor.Force 586 6573 1672 1920
## Retired 739 8000 1971 2402
## Unemployed 158 1782 466 554
## Prediction
## Reference Unemployed
## Disabled 169
## Employed 1843
## Not.in.Labor.Force 396
## Retired 501
## Unemployed 113
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.3982241 NA 0.3947671 0.4016888 0.5850800
## AccuracyPValue McnemarPValue
## 1.0000000 0.2051955
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 78 884 227 283
## Employed 891 9700 2466 2925
## Not.in.Labor.Force 236 2352 605 739
## Retired 283 2971 680 877
## Unemployed 72 676 143 201
## Prediction
## Reference Unemployed
## Disabled 64
## Employed 615
## Not.in.Labor.Force 167
## Retired 195
## Unemployed 38
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.3982656514 -0.0008038519 0.3925619972 0.4039901490 0.5850606317
## AccuracyPValue McnemarPValue
## 1.0000000000 0.3564451485
## model_id model_method feats max.nTuningRuns
## 1 Random.myrandom_classfr myrandom_classfr .rnorm 0
## min.elapsedtime.everything min.elapsedtime.final max.Accuracy.fit
## 1 0.304 0.01 0.3982241
## max.AccuracyLower.fit max.AccuracyUpper.fit max.Kappa.fit
## 1 0.3947671 0.4016888 NA
## max.Accuracy.OOB max.AccuracyLower.OOB max.AccuracyUpper.OOB
## 1 0.3982657 0.392562 0.4039901
## max.Kappa.OOB
## 1 -0.0008038519# Any models that have tuning parameters has "better" results with cross-validation
# (except rf) & "different" results for different outcome metrics
# Max.cor.Y
# Check impact of cv
# rpart is not a good candidate since caret does not optimize cp (only tuning parameter of rpart) well
ret_lst <- myfit_mdl(model_id="Max.cor.Y.cv.0",
model_method="rpart",
model_type=glb_model_type,
indep_vars_vctr=max_cor_y_x_vars,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df)## [1] "fitting model: Max.cor.Y.cv.0.rpart"
## [1] " indep_vars: Industry.my.fctr, Age"## Loading required package: rpart## Fitting cp = 0.252 on full training set## Loading required package: rpart.plot## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 0.2520541 0 1
##
## Node number 1: 77145 observations
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 4176 0 0
## Employed 0 45136 0 0
## Not.in.Labor.Force 0 11147 0 0
## Retired 0 13613 0 0
## Unemployed 0 3073 0 0
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.5850800 NA 0.5815936 0.5885601 0.5850800
## AccuracyPValue McnemarPValue
## 0.5015403 NaN
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 1536 0 0
## Employed 0 16597 0 0
## Not.in.Labor.Force 0 4099 0 0
## Retired 0 5006 0 0
## Unemployed 0 1130 0 0
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.5850606 0.0000000 0.5793009 0.5908029 0.5850606
## AccuracyPValue McnemarPValue
## 0.5025399 NaN
## model_id model_method feats max.nTuningRuns
## 1 Max.cor.Y.cv.0.rpart rpart Industry.my.fctr, Age 0
## min.elapsedtime.everything min.elapsedtime.final max.Accuracy.fit
## 1 13.775 4.064 0.58508
## max.AccuracyLower.fit max.AccuracyUpper.fit max.Kappa.fit
## 1 0.5815936 0.5885601 NA
## max.Accuracy.OOB max.AccuracyLower.OOB max.AccuracyUpper.OOB
## 1 0.5850606 0.5793009 0.5908029
## max.Kappa.OOB
## 1 0ret_lst <- myfit_mdl(model_id="Max.cor.Y.cv.0.cp.0",
model_method="rpart",
model_type=glb_model_type,
indep_vars_vctr=max_cor_y_x_vars,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=0,
tune_models_df=data.frame(parameter="cp", min=0.0, max=0.0, by=0.1))## [1] "fitting model: Max.cor.Y.cv.0.cp.0.rpart"
## [1] " indep_vars: Industry.my.fctr, Age"
## Fitting cp = 0 on full training set## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 2.520541e-01 0 1.0000000
## 2 9.075572e-02 1 0.7479459
## 3 1.930707e-02 2 0.6571902
## 4 1.190290e-02 3 0.6378831
## 5 1.134056e-02 4 0.6259802
## 6 1.065325e-02 5 0.6146396
## 7 9.256772e-03 6 0.6039864
## 8 7.591615e-03 19 0.4122903
## 9 6.310725e-03 20 0.4046987
## 10 6.248243e-03 22 0.3920772
## 11 6.217001e-03 23 0.3858290
## 12 6.092037e-03 24 0.3796120
## 13 6.029554e-03 25 0.3735199
## 14 4.717423e-03 26 0.3674904
## 15 4.592458e-03 27 0.3627730
## 16 4.092599e-03 28 0.3581805
## 17 3.842669e-03 29 0.3540879
## 18 3.811428e-03 30 0.3502452
## 19 3.249086e-03 31 0.3464338
## 20 3.238672e-03 32 0.3431847
## 21 2.936674e-03 41 0.2993221
## 22 2.530538e-03 42 0.2963854
## 23 1.749508e-03 43 0.2938549
## 24 1.687026e-03 44 0.2921053
## 25 1.499578e-03 45 0.2904183
## 26 1.030960e-03 46 0.2889187
## 27 9.684776e-04 47 0.2878878
## 28 9.059952e-04 49 0.2859508
## 29 7.497891e-04 50 0.2850448
## 30 6.248243e-04 51 0.2842950
## 31 4.998594e-04 52 0.2836702
## 32 4.373770e-04 53 0.2831704
## 33 3.748946e-04 54 0.2827330
## 34 2.030679e-04 55 0.2823581
## 35 1.718267e-04 57 0.2819520
## 36 6.248243e-05 59 0.2816083
## 37 3.124121e-05 60 0.2815458
## 38 0.000000e+00 63 0.2814521
##
## Variable importance
## Age
## 43
## Industry.my.fctrOther services
## 7
## Industry.my.fctrPublic administration
## 5
## Industry.my.fctrEducational and health services
## 5
## Industry.my.fctrLeisure and hospitality
## 5
## Industry.my.fctrTransportation and utilities
## 5
## Industry.my.fctrTrade
## 5
## Industry.my.fctrConstruction
## 4
## Industry.my.fctrFinancial
## 4
## Industry.my.fctrProfessional and business services
## 4
## Industry.my.fctrManufacturing
## 4
## Industry.my.fctrAgriculture, forestry, fishing, and hunting
## 4
## Industry.my.fctrInformation
## 3
## Industry.my.fctrMining
## 2
##
## Node number 1: 77145 observations, complexity param=0.2520541
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (62433 obs) right son=3 (14712 obs)
## Primary splits:
## Age < 64.5 to the left, improve=8861.8390, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=2323.6000, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=1186.7910, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 983.1942, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 924.0788, (0 missing)
##
## Node number 2: 62433 observations, complexity param=0.09075572
## predicted class=Employed expected loss=0.3215447 P(node) =0.8092942
## class counts: 3550 42358 10819 2767 2939
## probabilities: 0.057 0.678 0.173 0.044 0.047
## left son=4 (57429 obs) right son=5 (5004 obs)
## Primary splits:
## Age < 18.5 to the right, improve=3249.3410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1290.5850, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 624.4119, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 515.9783, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 505.1749, (0 missing)
##
## Node number 3: 14712 observations, complexity param=0.01930707
## predicted class=Retired expected loss=0.2627787 P(node) =0.1907058
## class counts: 626 2778 328 10846 134
## probabilities: 0.043 0.189 0.022 0.737 0.009
## left son=6 (696 obs) right son=7 (14016 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=756.4649, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=436.5917, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=408.2808, (0 missing)
## Age < 70.5 to the left, improve=364.4106, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=241.2709, (0 missing)
##
## Node number 4: 57429 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.2781347 P(node) =0.7444293
## class counts: 3490 41456 7012 2765 2706
## probabilities: 0.061 0.722 0.122 0.048 0.047
## left son=8 (10137 obs) right son=9 (47292 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=880.1351, (0 missing)
## Age < 59.5 to the left, improve=604.0624, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=395.1759, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=341.2702, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=334.6625, (0 missing)
##
## Node number 5: 5004 observations, complexity param=0.0119029
## predicted class=Not.in.Labor.Force expected loss=0.2392086 P(node) =0.06486486
## class counts: 60 902 3807 2 233
## probabilities: 0.012 0.180 0.761 0.000 0.047
## left son=10 (487 obs) right son=11 (4517 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=501.58970, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=243.57050, (0 missing)
## Age < 16.5 to the right, improve=107.27410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve= 66.39350, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 38.80304, (0 missing)
##
## Node number 6: 696 observations
## predicted class=Employed expected loss=0.07758621 P(node) =0.009021972
## class counts: 0 642 0 24 30
## probabilities: 0.000 0.922 0.000 0.034 0.043
##
## Node number 7: 14016 observations, complexity param=0.01134056
## predicted class=Retired expected loss=0.2278824 P(node) =0.1816838
## class counts: 626 2136 328 10822 104
## probabilities: 0.045 0.152 0.023 0.772 0.007
## left son=14 (402 obs) right son=15 (13614 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=481.2436, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=450.4300, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=265.4408, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=262.8831, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=260.1050, (0 missing)
##
## Node number 8: 10137 observations
## predicted class=Employed expected loss=0.05011345 P(node) =0.1314019
## class counts: 9 9629 51 13 435
## probabilities: 0.001 0.950 0.005 0.001 0.043
##
## Node number 9: 47292 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3270109 P(node) =0.6130274
## class counts: 3481 31827 6961 2752 2271
## probabilities: 0.074 0.673 0.147 0.058 0.048
## left son=18 (42249 obs) right son=19 (5043 obs)
## Primary splits:
## Age < 59.5 to the left, improve=682.6394, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=615.9417, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=526.5017, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=511.4015, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=354.9892, (0 missing)
##
## Node number 10: 487 observations
## predicted class=Employed expected loss=0.1457906 P(node) =0.006312788
## class counts: 0 416 35 0 36
## probabilities: 0.000 0.854 0.072 0.000 0.074
##
## Node number 11: 4517 observations, complexity param=0.006029554
## predicted class=Not.in.Labor.Force expected loss=0.1649325 P(node) =0.05855208
## class counts: 60 486 3772 2 197
## probabilities: 0.013 0.108 0.835 0.000 0.044
## left son=22 (232 obs) right son=23 (4285 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=298.35260, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve= 84.49187, (0 missing)
## Age < 16.5 to the right, improve= 55.14438, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 49.12110, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 48.50088, (0 missing)
##
## Node number 14: 402 observations
## predicted class=Employed expected loss=0.06716418 P(node) =0.005210966
## class counts: 2 375 0 12 13
## probabilities: 0.005 0.933 0.000 0.030 0.032
##
## Node number 15: 13614 observations, complexity param=0.01065325
## predicted class=Retired expected loss=0.2059644 P(node) =0.1764729
## class counts: 624 1761 328 10810 91
## probabilities: 0.046 0.129 0.024 0.794 0.007
## left son=30 (384 obs) right son=31 (13230 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=478.0754, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=281.2735, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=278.5598, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=275.8898, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=270.8294, (0 missing)
##
## Node number 18: 42249 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3028001 P(node) =0.547657
## class counts: 2844 29456 6676 1140 2133
## probabilities: 0.067 0.697 0.158 0.027 0.050
## left son=36 (5469 obs) right son=37 (36780 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=491.7069, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=427.5658, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=419.4708, (0 missing)
## Age < 23.5 to the right, improve=324.6411, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=289.1954, (0 missing)
##
## Node number 19: 5043 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.5298433 P(node) =0.06537041
## class counts: 637 2371 285 1612 138
## probabilities: 0.126 0.470 0.057 0.320 0.027
## left son=38 (451 obs) right son=39 (4592 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=169.22760, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=123.99940, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=118.31870, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 90.86328, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 76.67119, (0 missing)
##
## Node number 22: 232 observations
## predicted class=Employed expected loss=0.1336207 P(node) =0.003007324
## class counts: 0 201 8 0 23
## probabilities: 0.000 0.866 0.034 0.000 0.099
##
## Node number 23: 4285 observations, complexity param=0.001749508
## predicted class=Not.in.Labor.Force expected loss=0.1215869 P(node) =0.05554475
## class counts: 60 285 3764 2 174
## probabilities: 0.014 0.067 0.878 0.000 0.041
## left son=46 (97 obs) right son=47 (4188 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=95.90269, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=55.61024, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=55.21431, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=53.14430, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=39.60641, (0 missing)
##
## Node number 30: 384 observations
## predicted class=Employed expected loss=0.0859375 P(node) =0.00497764
## class counts: 1 351 2 10 20
## probabilities: 0.003 0.914 0.005 0.026 0.052
##
## Node number 31: 13230 observations, complexity param=0.006310725
## predicted class=Retired expected loss=0.1836735 P(node) =0.1714952
## class counts: 623 1410 326 10800 71
## probabilities: 0.047 0.107 0.025 0.816 0.005
## left son=62 (219 obs) right son=63 (13011 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=297.6126, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=294.7483, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=292.1900, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=286.7155, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=222.5232, (0 missing)
##
## Node number 36: 5469 observations
## predicted class=Employed expected loss=0.07899067 P(node) =0.07089248
## class counts: 7 5037 40 1 384
## probabilities: 0.001 0.921 0.007 0.000 0.070
##
## Node number 37: 36780 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3360794 P(node) =0.4767645
## class counts: 2837 24419 6636 1139 1749
## probabilities: 0.077 0.664 0.180 0.031 0.048
## left son=74 (4668 obs) right son=75 (32112 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=571.8024, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=556.1714, (0 missing)
## Age < 23.5 to the right, improve=394.3466, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=392.5422, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=371.2783, (0 missing)
##
## Node number 38: 451 observations
## predicted class=Employed expected loss=0.05543237 P(node) =0.005846134
## class counts: 1 426 1 3 20
## probabilities: 0.002 0.945 0.002 0.007 0.044
##
## Node number 39: 4592 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.5764373 P(node) =0.05952427
## class counts: 636 1945 284 1609 118
## probabilities: 0.139 0.424 0.062 0.350 0.026
## left son=78 (354 obs) right son=79 (4238 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=151.30200, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=145.26730, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=109.69160, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 93.33870, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve= 76.52206, (0 missing)
##
## Node number 46: 97 observations
## predicted class=Employed expected loss=0.2474227 P(node) =0.001257372
## class counts: 0 73 17 0 7
## probabilities: 0.000 0.753 0.175 0.000 0.072
##
## Node number 47: 4188 observations, complexity param=0.00103096
## predicted class=Not.in.Labor.Force expected loss=0.1053009 P(node) =0.05428738
## class counts: 60 212 3747 2 167
## probabilities: 0.014 0.051 0.895 0.000 0.040
## left son=94 (55 obs) right son=95 (4133 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=58.19129, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=57.88160, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=55.33508, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=41.20103, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=37.29178, (0 missing)
##
## Node number 62: 219 observations
## predicted class=Employed expected loss=0.05479452 P(node) =0.00283881
## class counts: 0 207 0 5 7
## probabilities: 0.000 0.945 0.000 0.023 0.032
##
## Node number 63: 13011 observations, complexity param=0.006248243
## predicted class=Retired expected loss=0.1703174 P(node) =0.1686564
## class counts: 623 1203 326 10795 64
## probabilities: 0.048 0.092 0.025 0.830 0.005
## left son=126 (217 obs) right son=127 (12794 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=304.8379, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=302.3581, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=296.6215, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=229.9711, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=198.7685, (0 missing)
##
## Node number 74: 4668 observations
## predicted class=Employed expected loss=0.07219366 P(node) =0.06050943
## class counts: 3 4331 31 0 303
## probabilities: 0.001 0.928 0.007 0.000 0.065
##
## Node number 75: 32112 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3744395 P(node) =0.4162551
## class counts: 2834 20088 6605 1139 1446
## probabilities: 0.088 0.626 0.206 0.035 0.045
## left son=150 (4364 obs) right son=151 (27748 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=739.5833, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=532.0889, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=481.7038, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=415.8067, (0 missing)
## Age < 23.5 to the right, improve=364.5501, (0 missing)
##
## Node number 78: 354 observations
## predicted class=Employed expected loss=0.07344633 P(node) =0.004588761
## class counts: 0 328 0 3 23
## probabilities: 0.000 0.927 0.000 0.008 0.065
##
## Node number 79: 4238 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.6184521 P(node) =0.05493551
## class counts: 636 1617 284 1606 95
## probabilities: 0.150 0.382 0.067 0.379 0.022
## left son=158 (361 obs) right son=159 (3877 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=172.59900, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=128.53290, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=110.07090, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve= 90.67191, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 88.92407, (0 missing)
##
## Node number 94: 55 observations
## predicted class=Employed expected loss=0.2363636 P(node) =0.0007129432
## class counts: 0 42 9 0 4
## probabilities: 0.000 0.764 0.164 0.000 0.073
##
## Node number 95: 4133 observations, complexity param=0.0009684776
## predicted class=Not.in.Labor.Force expected loss=0.09557222 P(node) =0.05357444
## class counts: 60 170 3738 2 163
## probabilities: 0.015 0.041 0.904 0.000 0.039
## left son=190 (61 obs) right son=191 (4072 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=59.50594, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=56.66486, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=42.16854, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=38.28705, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=27.80150, (0 missing)
##
## Node number 126: 217 observations
## predicted class=Employed expected loss=0.06912442 P(node) =0.002812885
## class counts: 0 202 1 2 12
## probabilities: 0.000 0.931 0.005 0.009 0.055
##
## Node number 127: 12794 observations, complexity param=0.006217001
## predicted class=Retired expected loss=0.1564014 P(node) =0.1658435
## class counts: 623 1001 325 10793 52
## probabilities: 0.049 0.078 0.025 0.844 0.004
## left son=254 (222 obs) right son=255 (12572 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=312.9565, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=306.9477, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=237.7246, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=205.6450, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=192.0610, (0 missing)
##
## Node number 150: 4364 observations
## predicted class=Employed expected loss=0.06393217 P(node) =0.0565688
## class counts: 2 4085 15 3 259
## probabilities: 0.000 0.936 0.003 0.001 0.059
##
## Node number 151: 27748 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.4232737 P(node) =0.3596863
## class counts: 2832 16003 6590 1136 1187
## probabilities: 0.102 0.577 0.237 0.041 0.043
## left son=302 (3747 obs) right son=303 (24001 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=743.1407, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=644.6047, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=571.9225, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=471.3980, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=458.7793, (0 missing)
##
## Node number 158: 361 observations
## predicted class=Employed expected loss=0.09418283 P(node) =0.0046795
## class counts: 1 327 0 5 28
## probabilities: 0.003 0.906 0.000 0.014 0.078
##
## Node number 159: 3877 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5870518 P(node) =0.05025601
## class counts: 635 1290 284 1601 67
## probabilities: 0.164 0.333 0.073 0.413 0.017
## left son=318 (242 obs) right son=319 (3635 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=152.6798, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=131.5716, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=108.9098, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=106.1004, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=102.9366, (0 missing)
##
## Node number 190: 61 observations
## predicted class=Employed expected loss=0.3606557 P(node) =0.0007907188
## class counts: 0 39 9 0 13
## probabilities: 0.000 0.639 0.148 0.000 0.213
##
## Node number 191: 4072 observations, complexity param=0.0009684776
## predicted class=Not.in.Labor.Force expected loss=0.08423379 P(node) =0.05278372
## class counts: 60 131 3729 2 150
## probabilities: 0.015 0.032 0.916 0.000 0.037
## left son=382 (42 obs) right son=383 (4030 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=58.11483, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=43.22076, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=39.39532, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=28.43424, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=24.96236, (0 missing)
##
## Node number 254: 222 observations
## predicted class=Employed expected loss=0.07207207 P(node) =0.002877698
## class counts: 0 206 1 7 8
## probabilities: 0.000 0.928 0.005 0.032 0.036
##
## Node number 255: 12572 observations, complexity param=0.006092037
## predicted class=Retired expected loss=0.1420617 P(node) =0.1629658
## class counts: 623 795 324 10786 44
## probabilities: 0.050 0.063 0.026 0.858 0.003
## left son=510 (215 obs) right son=511 (12357 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=317.9065, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=245.9525, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=212.9354, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=198.9882, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=198.2375, (0 missing)
##
## Node number 302: 3747 observations
## predicted class=Employed expected loss=0.09634374 P(node) =0.04857087
## class counts: 2 3386 43 3 313
## probabilities: 0.001 0.904 0.011 0.001 0.084
##
## Node number 303: 24001 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.4743136 P(node) =0.3111154
## class counts: 2830 12617 6547 1133 874
## probabilities: 0.118 0.526 0.273 0.047 0.036
## left son=606 (2696 obs) right son=607 (21305 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=849.2295, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=771.8555, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=616.0829, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=606.5147, (0 missing)
## Age < 26.5 to the right, improve=546.5604, (0 missing)
##
## Node number 318: 242 observations
## predicted class=Employed expected loss=0.04132231 P(node) =0.00313695
## class counts: 0 232 0 1 9
## probabilities: 0.000 0.959 0.000 0.004 0.037
##
## Node number 319: 3635 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5598349 P(node) =0.04711906
## class counts: 635 1058 284 1600 58
## probabilities: 0.175 0.291 0.078 0.440 0.016
## left son=638 (226 obs) right son=639 (3409 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=151.0187, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=125.3594, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=121.5617, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=117.7939, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=108.4572, (0 missing)
##
## Node number 382: 42 observations
## predicted class=Employed expected loss=0.1904762 P(node) =0.0005444293
## class counts: 0 34 2 0 6
## probabilities: 0.000 0.810 0.048 0.000 0.143
##
## Node number 383: 4030 observations, complexity param=0.0007497891
## predicted class=Not.in.Labor.Force expected loss=0.0751861 P(node) =0.05223929
## class counts: 60 97 3727 2 144
## probabilities: 0.015 0.024 0.925 0.000 0.036
## left son=766 (30 obs) right son=767 (4000 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=44.11017, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=40.32177, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=28.97328, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=25.41738, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=21.77548, (0 missing)
##
## Node number 510: 215 observations
## predicted class=Employed expected loss=0.06976744 P(node) =0.00278696
## class counts: 0 200 0 5 10
## probabilities: 0.000 0.930 0.000 0.023 0.047
##
## Node number 511: 12357 observations, complexity param=0.004717423
## predicted class=Retired expected loss=0.127539 P(node) =0.1601789
## class counts: 623 595 324 10781 34
## probabilities: 0.050 0.048 0.026 0.872 0.003
## left son=1022 (158 obs) right son=1023 (12199 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=254.39830, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=220.42630, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=206.10560, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=205.04780, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 80.04601, (0 missing)
##
## Node number 606: 2696 observations
## predicted class=Employed expected loss=0.03783383 P(node) =0.03494718
## class counts: 0 2594 7 1 94
## probabilities: 0.000 0.962 0.003 0.000 0.035
##
## Node number 607: 21305 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.5295471 P(node) =0.2761683
## class counts: 2830 10023 6540 1132 780
## probabilities: 0.133 0.470 0.307 0.053 0.037
## left son=1214 (2819 obs) right son=1215 (18486 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=1012.4120, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 789.2268, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 783.2167, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 680.7946, (0 missing)
## Age < 26.5 to the right, improve= 519.2679, (0 missing)
##
## Node number 638: 226 observations
## predicted class=Employed expected loss=0.0619469 P(node) =0.002929548
## class counts: 0 212 0 6 8
## probabilities: 0.000 0.938 0.000 0.027 0.035
##
## Node number 639: 3409 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5324142 P(node) =0.04418951
## class counts: 635 846 284 1594 50
## probabilities: 0.186 0.248 0.083 0.468 0.015
## left son=1278 (199 obs) right son=1279 (3210 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=143.50720, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=138.57380, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=134.13540, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=123.53620, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 76.56807, (0 missing)
##
## Node number 766: 30 observations
## predicted class=Employed expected loss=0.1666667 P(node) =0.0003888781
## class counts: 0 25 1 0 4
## probabilities: 0.000 0.833 0.033 0.000 0.133
##
## Node number 767: 4000 observations, complexity param=0.0006248243
## predicted class=Not.in.Labor.Force expected loss=0.0685 P(node) =0.05185041
## class counts: 60 72 3726 2 140
## probabilities: 0.015 0.018 0.931 0.001 0.035
## left son=1534 (36 obs) right son=1535 (3964 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=41.016160, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=29.377430, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=25.758690, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.067800, (0 missing)
## Age < 17.5 to the left, improve= 9.052118, (0 missing)
##
## Node number 1022: 158 observations
## predicted class=Employed expected loss=0.03164557 P(node) =0.002048091
## class counts: 0 153 0 2 3
## probabilities: 0.000 0.968 0.000 0.013 0.019
##
## Node number 1023: 12199 observations, complexity param=0.004092599
## predicted class=Retired expected loss=0.116403 P(node) =0.1581308
## class counts: 623 442 324 10779 31
## probabilities: 0.051 0.036 0.027 0.884 0.003
## left son=2046 (144 obs) right son=2047 (12055 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=226.33950, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=211.73010, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=210.42740, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 82.21373, (0 missing)
## Age < 69.5 to the left, improve= 54.47671, (0 missing)
##
## Node number 1214: 2819 observations
## predicted class=Employed expected loss=0.08300816 P(node) =0.03654158
## class counts: 2 2585 17 3 212
## probabilities: 0.001 0.917 0.006 0.001 0.075
##
## Node number 1215: 18486 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.5976415 P(node) =0.2396267
## class counts: 2828 7438 6523 1129 568
## probabilities: 0.153 0.402 0.353 0.061 0.031
## left son=2430 (2078 obs) right son=2431 (16408 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=1045.4260, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1044.0940, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 916.3636, (0 missing)
## Age < 26.5 to the right, improve= 504.6234, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 382.4998, (0 missing)
##
## Node number 1278: 199 observations
## predicted class=Employed expected loss=0.09547739 P(node) =0.002579558
## class counts: 1 180 0 3 15
## probabilities: 0.005 0.905 0.000 0.015 0.075
##
## Node number 1279: 3210 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5043614 P(node) =0.04160996
## class counts: 634 666 284 1591 35
## probabilities: 0.198 0.207 0.088 0.496 0.011
## left son=2558 (182 obs) right son=2559 (3028 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=156.29700, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=151.07500, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=139.20960, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 86.10358, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 47.69273, (0 missing)
##
## Node number 1534: 36 observations
## predicted class=Employed expected loss=0.3333333 P(node) =0.0004666537
## class counts: 0 24 4 0 8
## probabilities: 0.000 0.667 0.111 0.000 0.222
##
## Node number 1535: 3964 observations, complexity param=0.0004998594
## predicted class=Not.in.Labor.Force expected loss=0.06104945 P(node) =0.05138376
## class counts: 60 48 3722 2 132
## probabilities: 0.015 0.012 0.939 0.001 0.033
## left son=3070 (17 obs) right son=3071 (3947 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=29.804300, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=26.116770, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.374470, (0 missing)
## Age < 17.5 to the left, improve= 6.419004, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.571355, (0 missing)
##
## Node number 2046: 144 observations
## predicted class=Employed expected loss=0.09027778 P(node) =0.001866615
## class counts: 0 131 1 0 12
## probabilities: 0.000 0.910 0.007 0.000 0.083
##
## Node number 2047: 12055 observations, complexity param=0.003842669
## predicted class=Retired expected loss=0.1058482 P(node) =0.1562642
## class counts: 623 311 323 10779 19
## probabilities: 0.052 0.026 0.027 0.894 0.002
## left son=4094 (139 obs) right son=4095 (11916 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=216.92580, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=215.39200, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 84.21834, (0 missing)
## Age < 69.5 to the left, improve= 42.58446, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 21.11926, (0 missing)
##
## Node number 2430: 2078 observations
## predicted class=Employed expected loss=0.05293551 P(node) =0.02693629
## class counts: 0 1968 6 2 102
## probabilities: 0.000 0.947 0.003 0.001 0.049
##
## Node number 2431: 16408 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.6028157 P(node) =0.2126904
## class counts: 2828 5470 6517 1127 466
## probabilities: 0.172 0.333 0.397 0.069 0.028
## left son=4862 (1984 obs) right son=4863 (14424 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1338.0430, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=1188.9860, (0 missing)
## Age < 41.5 to the right, improve= 523.3124, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 491.6569, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 358.8742, (0 missing)
##
## Node number 2558: 182 observations
## predicted class=Employed expected loss=0.07142857 P(node) =0.002359194
## class counts: 0 169 0 1 12
## probabilities: 0.000 0.929 0.000 0.005 0.066
##
## Node number 2559: 3028 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.4749009 P(node) =0.03925076
## class counts: 634 497 284 1590 23
## probabilities: 0.209 0.164 0.094 0.525 0.008
## left son=5118 (174 obs) right son=5119 (2854 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=170.19160, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=156.88020, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 96.83817, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 53.76955, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 30.07375, (0 missing)
##
## Node number 3070: 17 observations
## predicted class=Employed expected loss=0.05882353 P(node) =0.0002203642
## class counts: 0 16 0 0 1
## probabilities: 0.000 0.941 0.000 0.000 0.059
##
## Node number 3071: 3947 observations, complexity param=0.000437377
## predicted class=Not.in.Labor.Force expected loss=0.05700532 P(node) =0.05116339
## class counts: 60 32 3722 2 131
## probabilities: 0.015 0.008 0.943 0.001 0.033
## left son=6142 (14 obs) right son=6143 (3933 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=26.335330, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.561670, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.647505, (0 missing)
## Age < 17.5 to the left, improve= 4.420919, (0 missing)
##
## Node number 4094: 139 observations
## predicted class=Employed expected loss=0.08633094 P(node) =0.001801802
## class counts: 0 127 1 4 7
## probabilities: 0.000 0.914 0.007 0.029 0.050
##
## Node number 4095: 11916 observations, complexity param=0.003811428
## predicted class=Retired expected loss=0.09575361 P(node) =0.1544624
## class counts: 623 184 322 10775 12
## probabilities: 0.052 0.015 0.027 0.904 0.001
## left son=8190 (128 obs) right son=8191 (11788 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=220.28360, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 86.19226, (0 missing)
## Age < 69.5 to the left, improve= 32.25519, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 21.65488, (0 missing)
##
## Node number 4862: 1984 observations
## predicted class=Employed expected loss=0.03175403 P(node) =0.0257178
## class counts: 1 1921 5 3 54
## probabilities: 0.001 0.968 0.003 0.002 0.027
##
## Node number 4863: 14424 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.5485302 P(node) =0.1869726
## class counts: 2827 3549 6512 1124 412
## probabilities: 0.196 0.246 0.451 0.078 0.029
## left son=9726 (1952 obs) right son=9727 (12472 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=1586.3880, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 648.4297, (0 missing)
## Age < 41.5 to the right, improve= 548.0990, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 472.6345, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 283.5885, (0 missing)
##
## Node number 5118: 174 observations
## predicted class=Employed expected loss=0.04597701 P(node) =0.002255493
## class counts: 0 166 1 5 2
## probabilities: 0.000 0.954 0.006 0.029 0.011
##
## Node number 5119: 2854 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.4446391 P(node) =0.03699527
## class counts: 634 331 283 1585 21
## probabilities: 0.222 0.116 0.099 0.555 0.007
## left son=10238 (162 obs) right son=10239 (2692 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=177.24140, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=109.19970, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 60.72953, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 33.66763, (0 missing)
## Age < 61.5 to the left, improve= 24.28196, (0 missing)
##
## Node number 6142: 14 observations
## predicted class=Employed expected loss=0 P(node) =0.0001814764
## class counts: 0 14 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 6143: 3933 observations, complexity param=0.0003748946
## predicted class=Not.in.Labor.Force expected loss=0.05364861 P(node) =0.05098192
## class counts: 60 18 3722 2 131
## probabilities: 0.015 0.005 0.946 0.001 0.033
## left son=12286 (12 obs) right son=12287 (3921 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=22.722820, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.713301, (0 missing)
## Age < 17.5 to the left, improve= 3.677531, (0 missing)
##
## Node number 8190: 128 observations
## predicted class=Employed expected loss=0.046875 P(node) =0.001659213
## class counts: 0 122 0 0 6
## probabilities: 0.000 0.953 0.000 0.000 0.047
##
## Node number 8191: 11788 observations, complexity param=0.001499578
## predicted class=Retired expected loss=0.08593485 P(node) =0.1528032
## class counts: 623 62 322 10775 6
## probabilities: 0.053 0.005 0.027 0.914 0.001
## left son=16382 (54 obs) right son=16383 (11734 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=88.14437, (0 missing)
## Age < 66.5 to the left, improve=24.62345, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=22.18490, (0 missing)
##
## Node number 9726: 1952 observations
## predicted class=Employed expected loss=0.07633197 P(node) =0.025303
## class counts: 2 1803 13 0 134
## probabilities: 0.001 0.924 0.007 0.000 0.069
##
## Node number 9727: 12472 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.4789128 P(node) =0.1616696
## class counts: 2825 1746 6499 1124 278
## probabilities: 0.227 0.140 0.521 0.090 0.022
## left son=19454 (864 obs) right son=19455 (11608 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=875.54450, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=636.69090, (0 missing)
## Age < 44.5 to the right, improve=628.14260, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=376.68880, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 25.43837, (0 missing)
##
## Node number 10238: 162 observations
## predicted class=Employed expected loss=0.0617284 P(node) =0.002099942
## class counts: 0 152 0 2 8
## probabilities: 0.000 0.938 0.000 0.012 0.049
##
## Node number 10239: 2692 observations, complexity param=0.002936674
## predicted class=Retired expected loss=0.4119614 P(node) =0.03489533
## class counts: 634 179 283 1583 13
## probabilities: 0.236 0.066 0.105 0.588 0.005
## left son=20478 (101 obs) right son=20479 (2591 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=122.93260, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 68.49246, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 37.63462, (0 missing)
## Age < 61.5 to the left, improve= 23.23422, (0 missing)
##
## Node number 12286: 12 observations
## predicted class=Employed expected loss=0 P(node) =0.0001555512
## class counts: 0 12 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 12287: 3921 observations, complexity param=6.248243e-05
## predicted class=Not.in.Labor.Force expected loss=0.05075236 P(node) =0.05082637
## class counts: 60 6 3722 2 131
## probabilities: 0.015 0.002 0.949 0.001 0.033
## left son=24574 (8 obs) right son=24575 (3913 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=5.770378, (0 missing)
## Age < 17.5 to the left, improve=3.474435, (0 missing)
##
## Node number 16382: 54 observations
## predicted class=Employed expected loss=0.09259259 P(node) =0.0006999806
## class counts: 0 49 0 1 4
## probabilities: 0.000 0.907 0.000 0.019 0.074
##
## Node number 16383: 11734 observations, complexity param=0.0001718267
## predicted class=Retired expected loss=0.08181353 P(node) =0.1521032
## class counts: 623 13 322 10774 2
## probabilities: 0.053 0.001 0.027 0.918 0.000
## left son=32766 (1428 obs) right son=32767 (10306 obs)
## Primary splits:
## Age < 66.5 to the left, improve=22.59645, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=22.40638, (0 missing)
##
## Node number 19454: 864 observations
## predicted class=Employed expected loss=0.07407407 P(node) =0.01119969
## class counts: 0 800 6 1 57
## probabilities: 0.000 0.926 0.007 0.001 0.066
##
## Node number 19455: 11608 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.4406444 P(node) =0.1504699
## class counts: 2825 946 6493 1123 221
## probabilities: 0.243 0.081 0.559 0.097 0.019
## left son=38910 (642 obs) right son=38911 (10966 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=738.10700, (0 missing)
## Age < 44.5 to the right, improve=681.98090, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=433.82320, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 26.34573, (0 missing)
##
## Node number 20478: 101 observations
## predicted class=Employed expected loss=0.04950495 P(node) =0.001309223
## class counts: 0 96 0 2 3
## probabilities: 0.000 0.950 0.000 0.020 0.030
##
## Node number 20479: 2591 observations, complexity param=0.001687026
## predicted class=Retired expected loss=0.3898109 P(node) =0.0335861
## class counts: 634 83 283 1581 10
## probabilities: 0.245 0.032 0.109 0.610 0.004
## left son=40958 (60 obs) right son=40959 (2531 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=74.12104, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=40.50551, (0 missing)
## Age < 61.5 to the left, improve=22.94488, (0 missing)
##
## Node number 24574: 8 observations
## predicted class=Employed expected loss=0.375 P(node) =0.0001037008
## class counts: 0 5 3 0 0
## probabilities: 0.000 0.625 0.375 0.000 0.000
##
## Node number 24575: 3913 observations
## predicted class=Not.in.Labor.Force expected loss=0.04957833 P(node) =0.05072267
## class counts: 60 1 3719 2 131
## probabilities: 0.015 0.000 0.950 0.001 0.033
##
## Node number 32766: 1428 observations
## predicted class=Retired expected loss=0.17507 P(node) =0.0185106
## class counts: 172 2 76 1178 0
## probabilities: 0.120 0.001 0.053 0.825 0.000
##
## Node number 32767: 10306 observations, complexity param=0.0001718267
## predicted class=Retired expected loss=0.06889191 P(node) =0.1335926
## class counts: 451 11 246 9596 2
## probabilities: 0.044 0.001 0.024 0.931 0.000
## left son=65534 (12 obs) right son=65535 (10294 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=20.600160, (0 missing)
## Age < 70.5 to the left, improve= 4.406275, (0 missing)
##
## Node number 38910: 642 observations
## predicted class=Employed expected loss=0.07165109 P(node) =0.008321991
## class counts: 2 596 7 1 36
## probabilities: 0.003 0.928 0.011 0.002 0.056
##
## Node number 38911: 10966 observations, complexity param=0.007591615
## predicted class=Not.in.Labor.Force expected loss=0.4085355 P(node) =0.1421479
## class counts: 2823 350 6486 1122 185
## probabilities: 0.257 0.032 0.591 0.102 0.017
## left son=77822 (4676 obs) right son=77823 (6290 obs)
## Primary splits:
## Age < 43.5 to the right, improve=710.94140, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=485.31870, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 27.25333, (0 missing)
##
## Node number 40958: 60 observations
## predicted class=Employed expected loss=0.1 P(node) =0.0007777562
## class counts: 0 54 0 0 6
## probabilities: 0.000 0.900 0.000 0.000 0.100
##
## Node number 40959: 2531 observations, complexity param=0.0009059952
## predicted class=Retired expected loss=0.3753457 P(node) =0.03280835
## class counts: 634 29 283 1581 4
## probabilities: 0.250 0.011 0.112 0.625 0.002
## left son=81918 (29 obs) right son=81919 (2502 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=42.32207, (0 missing)
## Age < 61.5 to the left, improve=22.59598, (0 missing)
##
## Node number 65534: 12 observations
## predicted class=Employed expected loss=0.08333333 P(node) =0.0001555512
## class counts: 0 11 0 0 1
## probabilities: 0.000 0.917 0.000 0.000 0.083
##
## Node number 65535: 10294 observations
## predicted class=Retired expected loss=0.06780649 P(node) =0.133437
## class counts: 451 0 246 9596 1
## probabilities: 0.044 0.000 0.024 0.932 0.000
##
## Node number 77822: 4676 observations, complexity param=0.004592458
## predicted class=Disabled expected loss=0.5902481 P(node) =0.06061313
## class counts: 1916 147 1673 927 13
## probabilities: 0.410 0.031 0.358 0.198 0.003
## left son=155644 (152 obs) right son=155645 (4524 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=190.3492, (0 missing)
## Age < 53.5 to the left, improve= 92.3401, (0 missing)
##
## Node number 77823: 6290 observations, complexity param=0.006310725
## predicted class=Not.in.Labor.Force expected loss=0.2348172 P(node) =0.08153477
## class counts: 907 203 4813 195 172
## probabilities: 0.144 0.032 0.765 0.031 0.027
## left son=155646 (211 obs) right son=155647 (6079 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=319.75960, (0 missing)
## Age < 26.5 to the right, improve= 58.87956, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 29.62228, (0 missing)
##
## Node number 81918: 29 observations
## predicted class=Employed expected loss=0 P(node) =0.0003759155
## class counts: 0 29 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 81919: 2502 observations
## predicted class=Retired expected loss=0.3681055 P(node) =0.03243243
## class counts: 634 0 283 1581 4
## probabilities: 0.253 0.000 0.113 0.632 0.002
##
## Node number 155644: 152 observations
## predicted class=Employed expected loss=0.03289474 P(node) =0.001970316
## class counts: 0 147 0 0 5
## probabilities: 0.000 0.967 0.000 0.000 0.033
##
## Node number 155645: 4524 observations, complexity param=0.003249086
## predicted class=Disabled expected loss=0.576481 P(node) =0.05864282
## class counts: 1916 0 1673 927 8
## probabilities: 0.424 0.000 0.370 0.205 0.002
## left son=311290 (2474 obs) right son=311291 (2050 obs)
## Primary splits:
## Age < 53.5 to the left, improve=96.05737, (0 missing)
##
## Node number 155646: 211 observations
## predicted class=Employed expected loss=0.03791469 P(node) =0.002735109
## class counts: 0 203 1 0 7
## probabilities: 0.000 0.962 0.005 0.000 0.033
##
## Node number 155647: 6079 observations, complexity param=0.0002030679
## predicted class=Not.in.Labor.Force expected loss=0.2084224 P(node) =0.07879966
## class counts: 907 0 4812 195 165
## probabilities: 0.149 0.000 0.792 0.032 0.027
## left son=311294 (3558 obs) right son=311295 (2521 obs)
## Primary splits:
## Age < 26.5 to the right, improve=51.65542, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve=30.42549, (0 missing)
##
## Node number 311290: 2474 observations, complexity param=0.002530538
## predicted class=Not.in.Labor.Force expected loss=0.5359741 P(node) =0.03206948
## class counts: 1044 0 1148 276 6
## probabilities: 0.422 0.000 0.464 0.112 0.002
## left son=622580 (1136 obs) right son=622581 (1338 obs)
## Primary splits:
## Age < 49.5 to the right, improve=20.27439, (0 missing)
##
## Node number 311291: 2050 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.5746341 P(node) =0.02657334
## class counts: 872 0 525 651 2
## probabilities: 0.425 0.000 0.256 0.318 0.001
## left son=622582 (987 obs) right son=622583 (1063 obs)
## Primary splits:
## Age < 56.5 to the left, improve=11.09287, (0 missing)
##
## Node number 311294: 3558 observations, complexity param=0.0002030679
## predicted class=Not.in.Labor.Force expected loss=0.2613828 P(node) =0.04612094
## class counts: 724 0 2628 153 53
## probabilities: 0.203 0.000 0.739 0.043 0.015
## left son=622588 (3545 obs) right son=622589 (13 obs)
## Primary splits:
## Industry.my.fctrArmed forces < 0.5 to the left, improve=20.34446, (0 missing)
## Age < 37.5 to the right, improve=13.39218, (0 missing)
##
## Node number 311295: 2521 observations
## predicted class=Not.in.Labor.Force expected loss=0.1336771 P(node) =0.03267872
## class counts: 183 0 2184 42 112
## probabilities: 0.073 0.000 0.866 0.017 0.044
##
## Node number 622580: 1136 observations
## predicted class=Disabled expected loss=0.5440141 P(node) =0.01472552
## class counts: 518 0 437 180 1
## probabilities: 0.456 0.000 0.385 0.158 0.001
##
## Node number 622581: 1338 observations
## predicted class=Not.in.Labor.Force expected loss=0.4686099 P(node) =0.01734396
## class counts: 526 0 711 96 5
## probabilities: 0.393 0.000 0.531 0.072 0.004
##
## Node number 622582: 987 observations
## predicted class=Disabled expected loss=0.5542047 P(node) =0.01279409
## class counts: 440 0 293 253 1
## probabilities: 0.446 0.000 0.297 0.256 0.001
##
## Node number 622583: 1063 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.593603 P(node) =0.01377925
## class counts: 432 0 232 398 1
## probabilities: 0.406 0.000 0.218 0.374 0.001
## left son=1245166 (314 obs) right son=1245167 (749 obs)
## Primary splits:
## Age < 57.5 to the left, improve=1.268934, (0 missing)
##
## Node number 622588: 3545 observations
## predicted class=Not.in.Labor.Force expected loss=0.2586742 P(node) =0.04595243
## class counts: 724 0 2628 153 40
## probabilities: 0.204 0.000 0.741 0.043 0.011
##
## Node number 622589: 13 observations
## predicted class=Unemployed expected loss=0 P(node) =0.0001685138
## class counts: 0 0 0 0 13
## probabilities: 0.000 0.000 0.000 0.000 1.000
##
## Node number 1245166: 314 observations
## predicted class=Disabled expected loss=0.5732484 P(node) =0.004070257
## class counts: 134 0 76 104 0
## probabilities: 0.427 0.000 0.242 0.331 0.000
##
## Node number 1245167: 749 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.6021362 P(node) =0.00970899
## class counts: 298 0 156 294 1
## probabilities: 0.398 0.000 0.208 0.393 0.001
## left son=2490334 (385 obs) right son=2490335 (364 obs)
## Primary splits:
## Age < 58.5 to the left, improve=0.1735246, (0 missing)
##
## Node number 2490334: 385 observations
## predicted class=Disabled expected loss=0.6 P(node) =0.004990602
## class counts: 154 0 84 147 0
## probabilities: 0.400 0.000 0.218 0.382 0.000
##
## Node number 2490335: 364 observations
## predicted class=Retired expected loss=0.5961538 P(node) =0.004718387
## class counts: 144 0 72 147 1
## probabilities: 0.396 0.000 0.198 0.404 0.003
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 64.5 62433 20075 Employed (0.057 0.68 0.17 0.044 0.047)
## 4) Age>=18.5 57429 15973 Employed (0.061 0.72 0.12 0.048 0.047)
## 8) Industry.my.fctrEducational and health services>=0.5 10137 508 Employed (0.00089 0.95 0.005 0.0013 0.043) *
## 9) Industry.my.fctrEducational and health services< 0.5 47292 15465 Employed (0.074 0.67 0.15 0.058 0.048)
## 18) Age< 59.5 42249 12793 Employed (0.067 0.7 0.16 0.027 0.05)
## 36) Industry.my.fctrTrade>=0.5 5469 432 Employed (0.0013 0.92 0.0073 0.00018 0.07) *
## 37) Industry.my.fctrTrade< 0.5 36780 12361 Employed (0.077 0.66 0.18 0.031 0.048)
## 74) Industry.my.fctrProfessional and business services>=0.5 4668 337 Employed (0.00064 0.93 0.0066 0 0.065) *
## 75) Industry.my.fctrProfessional and business services< 0.5 32112 12024 Employed (0.088 0.63 0.21 0.035 0.045)
## 150) Industry.my.fctrManufacturing>=0.5 4364 279 Employed (0.00046 0.94 0.0034 0.00069 0.059) *
## 151) Industry.my.fctrManufacturing< 0.5 27748 11745 Employed (0.1 0.58 0.24 0.041 0.043)
## 302) Industry.my.fctrLeisure and hospitality>=0.5 3747 361 Employed (0.00053 0.9 0.011 0.0008 0.084) *
## 303) Industry.my.fctrLeisure and hospitality< 0.5 24001 11384 Employed (0.12 0.53 0.27 0.047 0.036)
## 606) Industry.my.fctrFinancial>=0.5 2696 102 Employed (0 0.96 0.0026 0.00037 0.035) *
## 607) Industry.my.fctrFinancial< 0.5 21305 11282 Employed (0.13 0.47 0.31 0.053 0.037)
## 1214) Industry.my.fctrConstruction>=0.5 2819 234 Employed (0.00071 0.92 0.006 0.0011 0.075) *
## 1215) Industry.my.fctrConstruction< 0.5 18486 11048 Employed (0.15 0.4 0.35 0.061 0.031)
## 2430) Industry.my.fctrTransportation and utilities>=0.5 2078 110 Employed (0 0.95 0.0029 0.00096 0.049) *
## 2431) Industry.my.fctrTransportation and utilities< 0.5 16408 9891 Not.in.Labor.Force (0.17 0.33 0.4 0.069 0.028)
## 4862) Industry.my.fctrPublic administration>=0.5 1984 63 Employed (0.0005 0.97 0.0025 0.0015 0.027) *
## 4863) Industry.my.fctrPublic administration< 0.5 14424 7912 Not.in.Labor.Force (0.2 0.25 0.45 0.078 0.029)
## 9726) Industry.my.fctrOther services>=0.5 1952 149 Employed (0.001 0.92 0.0067 0 0.069) *
## 9727) Industry.my.fctrOther services< 0.5 12472 5973 Not.in.Labor.Force (0.23 0.14 0.52 0.09 0.022)
## 19454) Industry.my.fctrInformation>=0.5 864 64 Employed (0 0.93 0.0069 0.0012 0.066) *
## 19455) Industry.my.fctrInformation< 0.5 11608 5115 Not.in.Labor.Force (0.24 0.081 0.56 0.097 0.019)
## 38910) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 642 46 Employed (0.0031 0.93 0.011 0.0016 0.056) *
## 38911) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 10966 4480 Not.in.Labor.Force (0.26 0.032 0.59 0.1 0.017)
## 77822) Age>=43.5 4676 2760 Disabled (0.41 0.031 0.36 0.2 0.0028)
## 155644) Industry.my.fctrMining>=0.5 152 5 Employed (0 0.97 0 0 0.033) *
## 155645) Industry.my.fctrMining< 0.5 4524 2608 Disabled (0.42 0 0.37 0.2 0.0018)
## 311290) Age< 53.5 2474 1326 Not.in.Labor.Force (0.42 0 0.46 0.11 0.0024)
## 622580) Age>=49.5 1136 618 Disabled (0.46 0 0.38 0.16 0.00088) *
## 622581) Age< 49.5 1338 627 Not.in.Labor.Force (0.39 0 0.53 0.072 0.0037) *
## 311291) Age>=53.5 2050 1178 Disabled (0.43 0 0.26 0.32 0.00098)
## 622582) Age< 56.5 987 547 Disabled (0.45 0 0.3 0.26 0.001) *
## 622583) Age>=56.5 1063 631 Disabled (0.41 0 0.22 0.37 0.00094)
## 1245166) Age< 57.5 314 180 Disabled (0.43 0 0.24 0.33 0) *
## 1245167) Age>=57.5 749 451 Disabled (0.4 0 0.21 0.39 0.0013)
## 2490334) Age< 58.5 385 231 Disabled (0.4 0 0.22 0.38 0) *
## 2490335) Age>=58.5 364 217 Retired (0.4 0 0.2 0.4 0.0027) *
## 77823) Age< 43.5 6290 1477 Not.in.Labor.Force (0.14 0.032 0.77 0.031 0.027)
## 155646) Industry.my.fctrMining>=0.5 211 8 Employed (0 0.96 0.0047 0 0.033) *
## 155647) Industry.my.fctrMining< 0.5 6079 1267 Not.in.Labor.Force (0.15 0 0.79 0.032 0.027)
## 311294) Age>=26.5 3558 930 Not.in.Labor.Force (0.2 0 0.74 0.043 0.015)
## 622588) Industry.my.fctrArmed forces< 0.5 3545 917 Not.in.Labor.Force (0.2 0 0.74 0.043 0.011) *
## 622589) Industry.my.fctrArmed forces>=0.5 13 0 Unemployed (0 0 0 0 1) *
## 311295) Age< 26.5 2521 337 Not.in.Labor.Force (0.073 0 0.87 0.017 0.044) *
## 19) Age>=59.5 5043 2672 Employed (0.13 0.47 0.057 0.32 0.027)
## 38) Industry.my.fctrTrade>=0.5 451 25 Employed (0.0022 0.94 0.0022 0.0067 0.044) *
## 39) Industry.my.fctrTrade< 0.5 4592 2647 Employed (0.14 0.42 0.062 0.35 0.026)
## 78) Industry.my.fctrProfessional and business services>=0.5 354 26 Employed (0 0.93 0 0.0085 0.065) *
## 79) Industry.my.fctrProfessional and business services< 0.5 4238 2621 Employed (0.15 0.38 0.067 0.38 0.022)
## 158) Industry.my.fctrManufacturing>=0.5 361 34 Employed (0.0028 0.91 0 0.014 0.078) *
## 159) Industry.my.fctrManufacturing< 0.5 3877 2276 Retired (0.16 0.33 0.073 0.41 0.017)
## 318) Industry.my.fctrFinancial>=0.5 242 10 Employed (0 0.96 0 0.0041 0.037) *
## 319) Industry.my.fctrFinancial< 0.5 3635 2035 Retired (0.17 0.29 0.078 0.44 0.016)
## 638) Industry.my.fctrConstruction>=0.5 226 14 Employed (0 0.94 0 0.027 0.035) *
## 639) Industry.my.fctrConstruction< 0.5 3409 1815 Retired (0.19 0.25 0.083 0.47 0.015)
## 1278) Industry.my.fctrLeisure and hospitality>=0.5 199 19 Employed (0.005 0.9 0 0.015 0.075) *
## 1279) Industry.my.fctrLeisure and hospitality< 0.5 3210 1619 Retired (0.2 0.21 0.088 0.5 0.011)
## 2558) Industry.my.fctrTransportation and utilities>=0.5 182 13 Employed (0 0.93 0 0.0055 0.066) *
## 2559) Industry.my.fctrTransportation and utilities< 0.5 3028 1438 Retired (0.21 0.16 0.094 0.53 0.0076)
## 5118) Industry.my.fctrPublic administration>=0.5 174 8 Employed (0 0.95 0.0057 0.029 0.011) *
## 5119) Industry.my.fctrPublic administration< 0.5 2854 1269 Retired (0.22 0.12 0.099 0.56 0.0074)
## 10238) Industry.my.fctrOther services>=0.5 162 10 Employed (0 0.94 0 0.012 0.049) *
## 10239) Industry.my.fctrOther services< 0.5 2692 1109 Retired (0.24 0.066 0.11 0.59 0.0048)
## 20478) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 101 5 Employed (0 0.95 0 0.02 0.03) *
## 20479) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 2591 1010 Retired (0.24 0.032 0.11 0.61 0.0039)
## 40958) Industry.my.fctrInformation>=0.5 60 6 Employed (0 0.9 0 0 0.1) *
## 40959) Industry.my.fctrInformation< 0.5 2531 950 Retired (0.25 0.011 0.11 0.62 0.0016)
## 81918) Industry.my.fctrMining>=0.5 29 0 Employed (0 1 0 0 0) *
## 81919) Industry.my.fctrMining< 0.5 2502 921 Retired (0.25 0 0.11 0.63 0.0016) *
## 5) Age< 18.5 5004 1197 Not.in.Labor.Force (0.012 0.18 0.76 0.0004 0.047)
## 10) Industry.my.fctrLeisure and hospitality>=0.5 487 71 Employed (0 0.85 0.072 0 0.074) *
## 11) Industry.my.fctrLeisure and hospitality< 0.5 4517 745 Not.in.Labor.Force (0.013 0.11 0.84 0.00044 0.044)
## 22) Industry.my.fctrTrade>=0.5 232 31 Employed (0 0.87 0.034 0 0.099) *
## 23) Industry.my.fctrTrade< 0.5 4285 521 Not.in.Labor.Force (0.014 0.067 0.88 0.00047 0.041)
## 46) Industry.my.fctrEducational and health services>=0.5 97 24 Employed (0 0.75 0.18 0 0.072) *
## 47) Industry.my.fctrEducational and health services< 0.5 4188 441 Not.in.Labor.Force (0.014 0.051 0.89 0.00048 0.04)
## 94) Industry.my.fctrOther services>=0.5 55 13 Employed (0 0.76 0.16 0 0.073) *
## 95) Industry.my.fctrOther services< 0.5 4133 395 Not.in.Labor.Force (0.015 0.041 0.9 0.00048 0.039)
## 190) Industry.my.fctrProfessional and business services>=0.5 61 22 Employed (0 0.64 0.15 0 0.21) *
## 191) Industry.my.fctrProfessional and business services< 0.5 4072 343 Not.in.Labor.Force (0.015 0.032 0.92 0.00049 0.037)
## 382) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 42 8 Employed (0 0.81 0.048 0 0.14) *
## 383) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 4030 303 Not.in.Labor.Force (0.015 0.024 0.92 0.0005 0.036)
## 766) Industry.my.fctrConstruction>=0.5 30 5 Employed (0 0.83 0.033 0 0.13) *
## 767) Industry.my.fctrConstruction< 0.5 4000 274 Not.in.Labor.Force (0.015 0.018 0.93 0.0005 0.035)
## 1534) Industry.my.fctrManufacturing>=0.5 36 12 Employed (0 0.67 0.11 0 0.22) *
## 1535) Industry.my.fctrManufacturing< 0.5 3964 242 Not.in.Labor.Force (0.015 0.012 0.94 0.0005 0.033)
## 3070) Industry.my.fctrTransportation and utilities>=0.5 17 1 Employed (0 0.94 0 0 0.059) *
## 3071) Industry.my.fctrTransportation and utilities< 0.5 3947 225 Not.in.Labor.Force (0.015 0.0081 0.94 0.00051 0.033)
## 6142) Industry.my.fctrFinancial>=0.5 14 0 Employed (0 1 0 0 0) *
## 6143) Industry.my.fctrFinancial< 0.5 3933 211 Not.in.Labor.Force (0.015 0.0046 0.95 0.00051 0.033)
## 12286) Industry.my.fctrInformation>=0.5 12 0 Employed (0 1 0 0 0) *
## 12287) Industry.my.fctrInformation< 0.5 3921 199 Not.in.Labor.Force (0.015 0.0015 0.95 0.00051 0.033)
## 24574) Industry.my.fctrPublic administration>=0.5 8 3 Employed (0 0.63 0.37 0 0) *
## 24575) Industry.my.fctrPublic administration< 0.5 3913 194 Not.in.Labor.Force (0.015 0.00026 0.95 0.00051 0.033) *
## 3) Age>=64.5 14712 3866 Retired (0.043 0.19 0.022 0.74 0.0091)
## 6) Industry.my.fctrEducational and health services>=0.5 696 54 Employed (0 0.92 0 0.034 0.043) *
## 7) Industry.my.fctrEducational and health services< 0.5 14016 3194 Retired (0.045 0.15 0.023 0.77 0.0074)
## 14) Industry.my.fctrTrade>=0.5 402 27 Employed (0.005 0.93 0 0.03 0.032) *
## 15) Industry.my.fctrTrade< 0.5 13614 2804 Retired (0.046 0.13 0.024 0.79 0.0067)
## 30) Industry.my.fctrProfessional and business services>=0.5 384 33 Employed (0.0026 0.91 0.0052 0.026 0.052) *
## 31) Industry.my.fctrProfessional and business services< 0.5 13230 2430 Retired (0.047 0.11 0.025 0.82 0.0054)
## 62) Industry.my.fctrFinancial>=0.5 219 12 Employed (0 0.95 0 0.023 0.032) *
## 63) Industry.my.fctrFinancial< 0.5 13011 2216 Retired (0.048 0.092 0.025 0.83 0.0049)
## 126) Industry.my.fctrManufacturing>=0.5 217 15 Employed (0 0.93 0.0046 0.0092 0.055) *
## 127) Industry.my.fctrManufacturing< 0.5 12794 2001 Retired (0.049 0.078 0.025 0.84 0.0041)
## 254) Industry.my.fctrOther services>=0.5 222 16 Employed (0 0.93 0.0045 0.032 0.036) *
## 255) Industry.my.fctrOther services< 0.5 12572 1786 Retired (0.05 0.063 0.026 0.86 0.0035)
## 510) Industry.my.fctrLeisure and hospitality>=0.5 215 15 Employed (0 0.93 0 0.023 0.047) *
## 511) Industry.my.fctrLeisure and hospitality< 0.5 12357 1576 Retired (0.05 0.048 0.026 0.87 0.0028)
## 1022) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 158 5 Employed (0 0.97 0 0.013 0.019) *
## 1023) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 12199 1420 Retired (0.051 0.036 0.027 0.88 0.0025)
## 2046) Industry.my.fctrConstruction>=0.5 144 13 Employed (0 0.91 0.0069 0 0.083) *
## 2047) Industry.my.fctrConstruction< 0.5 12055 1276 Retired (0.052 0.026 0.027 0.89 0.0016)
## 4094) Industry.my.fctrTransportation and utilities>=0.5 139 12 Employed (0 0.91 0.0072 0.029 0.05) *
## 4095) Industry.my.fctrTransportation and utilities< 0.5 11916 1141 Retired (0.052 0.015 0.027 0.9 0.001)
## 8190) Industry.my.fctrPublic administration>=0.5 128 6 Employed (0 0.95 0 0 0.047) *
## 8191) Industry.my.fctrPublic administration< 0.5 11788 1013 Retired (0.053 0.0053 0.027 0.91 0.00051)
## 16382) Industry.my.fctrInformation>=0.5 54 5 Employed (0 0.91 0 0.019 0.074) *
## 16383) Industry.my.fctrInformation< 0.5 11734 960 Retired (0.053 0.0011 0.027 0.92 0.00017)
## 32766) Age< 66.5 1428 250 Retired (0.12 0.0014 0.053 0.82 0) *
## 32767) Age>=66.5 10306 710 Retired (0.044 0.0011 0.024 0.93 0.00019)
## 65534) Industry.my.fctrMining>=0.5 12 1 Employed (0 0.92 0 0 0.083) *
## 65535) Industry.my.fctrMining< 0.5 10294 698 Retired (0.044 0 0.024 0.93 9.7e-05) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 1246 36 1493 1401
## Employed 0 45133 1 2
## Not.in.Labor.Force 890 338 9242 677
## Retired 684 134 293 12502
## Unemployed 2 2764 288 6
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 13
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.8832199 NA 0.8809327 0.8854782 0.5850800
## AccuracyPValue McnemarPValue
## 0.0000000 0.0000000
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 468 24 534 510
## Employed 0 16594 1 2
## Not.in.Labor.Force 348 133 3369 248
## Retired 244 41 111 4610
## Unemployed 3 1006 117 1
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 1
## Retired 0
## Unemployed 3
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 8.828257e-01 7.967286e-01 8.790261e-01 8.865467e-01 5.850606e-01
## AccuracyPValue McnemarPValue
## 0.000000e+00 4.618022e-316
## model_id model_method feats
## 1 Max.cor.Y.cv.0.cp.0.rpart rpart Industry.my.fctr, Age
## max.nTuningRuns min.elapsedtime.everything min.elapsedtime.final
## 1 0 4.443 3.816
## max.Accuracy.fit max.AccuracyLower.fit max.AccuracyUpper.fit
## 1 0.8832199 0.8809327 0.8854782
## max.Kappa.fit max.Accuracy.OOB max.AccuracyLower.OOB
## 1 NA 0.8828257 0.8790261
## max.AccuracyUpper.OOB max.Kappa.OOB
## 1 0.8865467 0.7967286if (glb_is_regression || glb_is_binomial) # For multinomials this model will be run next by default
ret_lst <- myfit_mdl(model_id="Max.cor.Y",
model_method="rpart",
model_type=glb_model_type,
indep_vars_vctr=max_cor_y_x_vars,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)
# Used to compare vs. Interactions.High.cor.Y and/or Max.cor.Y.TmSrs
ret_lst <- myfit_mdl(model_id="Max.cor.Y",
model_method=ifelse(glb_is_regression, "lm",
ifelse(glb_is_binomial, "glm", "rpart")),
model_type=glb_model_type,
indep_vars_vctr=max_cor_y_x_vars,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)## [1] "fitting model: Max.cor.Y.rpart"
## [1] " indep_vars: Industry.my.fctr, Age"
## Aggregating results
## Selecting tuning parameters
## Fitting cp = 0.0193 on full training set## Warning in myfit_mdl(model_id = "Max.cor.Y", model_method =
## ifelse(glb_is_regression, : model's bestTune found at an extreme of
## tuneGrid for parameter: cp
## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 0.25205411 0 1.0000000
## 2 0.09075572 1 0.7479459
## 3 0.01930707 2 0.6571902
##
## Variable importance
## Age
## 100
##
## Node number 1: 77145 observations, complexity param=0.2520541
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (62433 obs) right son=3 (14712 obs)
## Primary splits:
## Age < 64.5 to the left, improve=8861.8390, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=2323.6000, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=1186.7910, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 983.1942, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 924.0788, (0 missing)
##
## Node number 2: 62433 observations, complexity param=0.09075572
## predicted class=Employed expected loss=0.3215447 P(node) =0.8092942
## class counts: 3550 42358 10819 2767 2939
## probabilities: 0.057 0.678 0.173 0.044 0.047
## left son=4 (57429 obs) right son=5 (5004 obs)
## Primary splits:
## Age < 18.5 to the right, improve=3249.3410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1290.5850, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 624.4119, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 515.9783, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 505.1749, (0 missing)
##
## Node number 3: 14712 observations
## predicted class=Retired expected loss=0.2627787 P(node) =0.1907058
## class counts: 626 2778 328 10846 134
## probabilities: 0.043 0.189 0.022 0.737 0.009
##
## Node number 4: 57429 observations
## predicted class=Employed expected loss=0.2781347 P(node) =0.7444293
## class counts: 3490 41456 7012 2765 2706
## probabilities: 0.061 0.722 0.122 0.048 0.047
##
## Node number 5: 5004 observations
## predicted class=Not.in.Labor.Force expected loss=0.2392086 P(node) =0.06486486
## class counts: 60 902 3807 2 233
## probabilities: 0.012 0.180 0.761 0.000 0.047
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 64.5 62433 20075 Employed (0.057 0.68 0.17 0.044 0.047)
## 4) Age>=18.5 57429 15973 Employed (0.061 0.72 0.12 0.048 0.047) *
## 5) Age< 18.5 5004 1197 Not.in.Labor.Force (0.012 0.18 0.76 0.0004 0.047) *
## 3) Age>=64.5 14712 3866 Retired (0.043 0.19 0.022 0.74 0.0091) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 3490 60 626
## Employed 0 41456 902 2778
## Not.in.Labor.Force 0 7012 3807 328
## Retired 0 2765 2 10846
## Unemployed 0 2706 233 134
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7273187 NA 0.7241611 0.7304591 0.5850800
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 1289 21 226
## Employed 0 15206 365 1026
## Not.in.Labor.Force 0 2581 1416 102
## Retired 0 1028 1 3977
## Unemployed 0 986 97 47
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7261351 0.4752624 0.7209052 0.7313187 0.5850606
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## model_id model_method feats max.nTuningRuns
## 1 Max.cor.Y.rpart rpart Industry.my.fctr, Age 3
## min.elapsedtime.everything min.elapsedtime.final max.Accuracy.fit
## 1 17.853 3.83 0.73205
## max.AccuracyLower.fit max.AccuracyUpper.fit max.Kappa.fit
## 1 0.7241611 0.7304591 0.4874172
## max.Accuracy.OOB max.AccuracyLower.OOB max.AccuracyUpper.OOB
## 1 0.7261351 0.7209052 0.7313187
## max.Kappa.OOB max.AccuracySD.fit max.KappaSD.fit
## 1 0.4752624 0.004250054 0.002519487if (!is.null(glb_date_vars) &&
(sum(grepl(paste(glb_date_vars, "\\.day\\.minutes\\.poly\\.", sep=""),
names(glb_allobs_df))) > 0)) {
# ret_lst <- myfit_mdl(model_id="Max.cor.Y.TmSrs.poly1",
# model_method=ifelse(glb_is_regression, "lm",
# ifelse(glb_is_binomial, "glm", "rpart")),
# model_type=glb_model_type,
# indep_vars_vctr=c(max_cor_y_x_vars, paste0(glb_date_vars, ".day.minutes")),
# rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
# fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
# n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)
#
ret_lst <- myfit_mdl(model_id="Max.cor.Y.TmSrs.poly",
model_method=ifelse(glb_is_regression, "lm",
ifelse(glb_is_binomial, "glm", "rpart")),
model_type=glb_model_type,
indep_vars_vctr=c(max_cor_y_x_vars,
grep(paste(glb_date_vars, "\\.day\\.minutes\\.poly\\.", sep=""),
names(glb_allobs_df), value=TRUE)),
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)
}
# Interactions.High.cor.Y
if (length(int_feats <- setdiff(unique(glb_feats_df$cor.high.X), NA)) > 0) {
# lm & glm handle interaction terms; rpart & rf do not
if (glb_is_regression || glb_is_binomial) {
indep_vars_vctr <-
c(max_cor_y_x_vars, paste(max_cor_y_x_vars[1], int_feats, sep=":"))
} else { indep_vars_vctr <- union(max_cor_y_x_vars, int_feats) }
ret_lst <- myfit_mdl(model_id="Interact.High.cor.Y",
model_method=ifelse(glb_is_regression, "lm",
ifelse(glb_is_binomial, "glm", "rpart")),
model_type=glb_model_type,
indep_vars_vctr,
glb_rsp_var, glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)
}
# Low.cor.X
# if (glb_is_classification && glb_is_binomial)
# indep_vars_vctr <- subset(glb_feats_df, is.na(cor.high.X) &
# is.ConditionalX.y &
# (exclude.as.feat != 1))[, "id"] else
indep_vars_vctr <- subset(glb_feats_df, is.na(cor.high.X) & !myNearZV &
(exclude.as.feat != 1))[, "id"]
myadjust_interaction_feats <- function(vars_vctr) {
for (feat in subset(glb_feats_df, !is.na(interaction.feat))$id)
if (feat %in% vars_vctr)
vars_vctr <- union(setdiff(vars_vctr, feat),
paste0(glb_feats_df[glb_feats_df$id == feat, "interaction.feat"], ":", feat))
return(vars_vctr)
}
indep_vars_vctr <- myadjust_interaction_feats(indep_vars_vctr)
ret_lst <- myfit_mdl(model_id="Low.cor.X",
model_method=ifelse(glb_is_regression, "lm",
ifelse(glb_is_binomial, "glm", "rpart")),
indep_vars_vctr=indep_vars_vctr,
model_type=glb_model_type,
glb_rsp_var, glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=NULL)## [1] "fitting model: Low.cor.X.rpart"
## [1] " indep_vars: Age, Sex.fctr, Married.my.fctr, Region.fctr, .rnorm, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr"
## Aggregating results
## Selecting tuning parameters
## Fitting cp = 0.0193 on full training set## Warning in myfit_mdl(model_id = "Low.cor.X", model_method =
## ifelse(glb_is_regression, : model's bestTune found at an extreme of
## tuneGrid for parameter: cp
## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 0.25205411 0 1.0000000
## 2 0.09075572 1 0.7479459
## 3 0.01930707 2 0.6571902
##
## Variable importance
## Age Married.my.fctrWidowed
## 88 12
##
## Node number 1: 77145 observations, complexity param=0.2520541
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (62433 obs) right son=3 (14712 obs)
## Primary splits:
## Age < 64.5 to the left, improve=8861.839, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=2323.600, (0 missing)
## Married.my.fctrWidowed < 0.5 to the left, improve=2200.537, (0 missing)
## PeopleInHousehold < 2.5 to the right, improve=1947.586, (0 missing)
## Married.my.fctrNever Married < 0.5 to the right, improve=1420.837, (0 missing)
## Surrogate splits:
## Married.my.fctrWidowed < 0.5 to the left, agree=0.843, adj=0.178, (0 split)
## .rnorm < 3.985793 to the left, agree=0.809, adj=0.000, (0 split)
##
## Node number 2: 62433 observations, complexity param=0.09075572
## predicted class=Employed expected loss=0.3215447 P(node) =0.8092942
## class counts: 3550 42358 10819 2767 2939
## probabilities: 0.057 0.678 0.173 0.044 0.047
## left son=4 (57429 obs) right son=5 (5004 obs)
## Primary splits:
## Age < 18.5 to the right, improve=3249.3410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1290.5850, (0 missing)
## Married.my.fctrNever Married < 0.5 to the left, improve=1036.8480, (0 missing)
## Married.my.fctrMarried < 0.5 to the right, improve= 695.1223, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 624.4119, (0 missing)
## Surrogate splits:
## .rnorm < 3.928213 to the left, agree=0.92, adj=0, (0 split)
##
## Node number 3: 14712 observations
## predicted class=Retired expected loss=0.2627787 P(node) =0.1907058
## class counts: 626 2778 328 10846 134
## probabilities: 0.043 0.189 0.022 0.737 0.009
##
## Node number 4: 57429 observations
## predicted class=Employed expected loss=0.2781347 P(node) =0.7444293
## class counts: 3490 41456 7012 2765 2706
## probabilities: 0.061 0.722 0.122 0.048 0.047
##
## Node number 5: 5004 observations
## predicted class=Not.in.Labor.Force expected loss=0.2392086 P(node) =0.06486486
## class counts: 60 902 3807 2 233
## probabilities: 0.012 0.180 0.761 0.000 0.047
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 64.5 62433 20075 Employed (0.057 0.68 0.17 0.044 0.047)
## 4) Age>=18.5 57429 15973 Employed (0.061 0.72 0.12 0.048 0.047) *
## 5) Age< 18.5 5004 1197 Not.in.Labor.Force (0.012 0.18 0.76 0.0004 0.047) *
## 3) Age>=64.5 14712 3866 Retired (0.043 0.19 0.022 0.74 0.0091) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 3490 60 626
## Employed 0 41456 902 2778
## Not.in.Labor.Force 0 7012 3807 328
## Retired 0 2765 2 10846
## Unemployed 0 2706 233 134
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7273187 NA 0.7241611 0.7304591 0.5850800
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 1289 21 226
## Employed 0 15206 365 1026
## Not.in.Labor.Force 0 2581 1416 102
## Retired 0 1028 1 3977
## Unemployed 0 986 97 47
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7261351 0.4752624 0.7209052 0.7313187 0.5850606
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## model_id model_method
## 1 Low.cor.X.rpart rpart
## feats
## 1 Age, Sex.fctr, Married.my.fctr, Region.fctr, .rnorm, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## max.nTuningRuns min.elapsedtime.everything min.elapsedtime.final
## 1 3 57.322 10.976
## max.Accuracy.fit max.AccuracyLower.fit max.AccuracyUpper.fit
## 1 0.73205 0.7241611 0.7304591
## max.Kappa.fit max.Accuracy.OOB max.AccuracyLower.OOB
## 1 0.4874172 0.7261351 0.7209052
## max.AccuracyUpper.OOB max.Kappa.OOB max.AccuracySD.fit max.KappaSD.fit
## 1 0.7313187 0.4752624 0.004250054 0.002519487rm(ret_lst)
glb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.models", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 10 fit.models 7 0 75.172 186.411 111.239
## 11 fit.models 7 1 186.412 NA NAfit.models_1_chunk_df <- myadd_chunk(NULL, "fit.models_1_bgn")## label step_major step_minor bgn end elapsed
## 1 fit.models_1_bgn 1 0 190.32 NA NA# Options:
# 1. rpart & rf manual tuning
# 2. rf without pca (default: with pca)
#stop(here); sav_models_lst <- glb_models_lst; sav_models_df <- glb_models_df
#glb_models_lst <- sav_models_lst; glb_models_df <- sav_models_df
# All X that is not user excluded
# if (glb_is_classification && glb_is_binomial) {
# model_id_pfx <- "Conditional.X"
# # indep_vars_vctr <- setdiff(names(glb_fitobs_df), union(glb_rsp_var, glb_exclude_vars_as_features))
# indep_vars_vctr <- subset(glb_feats_df, is.ConditionalX.y &
# (exclude.as.feat != 1))[, "id"]
# } else {
model_id_pfx <- "All.X"
indep_vars_vctr <- subset(glb_feats_df, !myNearZV &
(exclude.as.feat != 1))[, "id"]
# }
indep_vars_vctr <- myadjust_interaction_feats(indep_vars_vctr)
for (method in glb_models_method_vctr) {
fit.models_1_chunk_df <- myadd_chunk(fit.models_1_chunk_df,
paste0("fit.models_1_", method), major.inc=TRUE)
if (method %in% c("rpart", "rf")) {
# rpart: fubar's the tree
# rf: skip the scenario w/ .rnorm for speed
indep_vars_vctr <- setdiff(indep_vars_vctr, c(".rnorm"))
model_id <- paste0(model_id_pfx, ".no.rnorm")
} else model_id <- model_id_pfx
if (method %in% c("glm")) # for a "robust" glm model
indep_vars_vctr <- setdiff(indep_vars_vctr, c(NULL
,"A.nchrs.log" # correlated to "S.*"
,"A.ndgts.log" # correlated to "S.*"
,"A.nuppr.log" # correlated to "S.*"
,"A.npnct01.log" # identical to "S.npnct01.log"
,"A.npnct03.log" # correlated to "S.npnct03.log"
,"A.npnct04.log" # correlated to "S.npnct04.log"
,"A.npnct06.log" # identical to "S.npnct06.log"
,"A.npnct07.log" # identical to "S.npnct07.log"
,"A.npnct08.log" # correlated to "S.npnct08.log"
,"A.npnct11.log" # correlated to "S.*"
,"A.npnct12.log" # correlated to "S.*"
,"S.npnct14.log" # correlated to "A.*"
,"A.npnct15.log" # correlated to "S.npnct15.log"
,"A.npnct16.log" # correlated to "S.npnct16.log"
,"A.npnct19.log" # correlated to "S.*"
,"A.npnct20.log" # identical to "S.npnct20.log"
,"A.npnct21.log" # correlated to "S.npnct21.log"
,"A.P.daily.clip.report" # identical to "S.*"
,"S.P.daily.clip.report" # identical to "H.*"
,"A.P.http" # correlated to "A.npnct14.log"
,"A.P.fashion.week" # identical to "S.*"
,"H.P.first.draft" # correlated to "H.T.first"
,"A.P.first.draft" # identical to "S.*"
,"A.P.metropolitan.diary.colon" # identical to "S.*"
,"A.P.year.colon" # identical to "S.P.year.colon"
))
ret_lst <- myfit_mdl(model_id=model_id, model_method=method,
indep_vars_vctr=indep_vars_vctr,
model_type=glb_model_type,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
n_cv_folds=glb_n_cv_folds, tune_models_df=glb_tune_models_df)
# If All.X.glm is less accurate than Low.Cor.X.glm
# check NA coefficients & filter appropriate terms in indep_vars_vctr
# if (method == "glm") {
# orig_glm <- glb_models_lst[[paste0(model_id, ".", model_method)]]$finalModel
# orig_glm <- glb_models_lst[["All.X.glm"]]$finalModel; print(summary(orig_glm))
# vif_orig_glm <- vif(orig_glm); print(vif_orig_glm)
# print(vif_orig_glm[!is.na(vif_orig_glm) & (vif_orig_glm == Inf)])
# print(which.max(vif_orig_glm))
# print(sort(vif_orig_glm[vif_orig_glm >= 1.0e+03], decreasing=TRUE))
# glb_fitobs_df[c(1143, 3637, 3953, 4105), c("UniqueID", "Popular", "H.P.quandary", "Headline")]
# glb_feats_df[glb_feats_df$id %in% grep("[HSA]\\.nchrs.log", glb_feats_df$id, value=TRUE) | glb_feats_df$cor.high.X %in% grep("[HSA]\\.nchrs.log", glb_feats_df$id, value=TRUE), ]
# glb_feats_df[glb_feats_df$id %in% grep("[HSA]\\.npnct14.log", glb_feats_df$id, value=TRUE) | glb_feats_df$cor.high.X %in% grep("[HSA]\\.npnct14.log", glb_feats_df$id, value=TRUE), ]
# glb_feats_df[glb_feats_df$id %in% grep("[HSA]\\.T.scen", glb_feats_df$id, value=TRUE) | glb_feats_df$cor.high.X %in% grep("[HSA]\\.T.scen", glb_feats_df$id, value=TRUE), ]
# glb_feats_df[glb_feats_df$id %in% grep("[HSA]\\.P.first", glb_feats_df$id, value=TRUE) | glb_feats_df$cor.high.X %in% grep("[HSA]\\.P.first", glb_feats_df$id, value=TRUE), ]
# all.equal(glb_allobs_df$S.nuppr.log, glb_allobs_df$A.nuppr.log)
# all.equal(glb_allobs_df$S.npnct19.log, glb_allobs_df$A.npnct19.log)
# all.equal(glb_allobs_df$S.P.year.colon, glb_allobs_df$A.P.year.colon)
# all.equal(glb_allobs_df$S.T.share, glb_allobs_df$A.T.share)
# all.equal(glb_allobs_df$H.T.clip, glb_allobs_df$H.P.daily.clip.report)
# cor(glb_allobs_df$S.T.herald, glb_allobs_df$S.T.tribun)
# dsp_obs(Abstract.contains="[Dd]iar", cols=("Abstract"), all=TRUE)
# dsp_obs(Abstract.contains="[Ss]hare", cols=("Abstract"), all=TRUE)
# subset(glb_feats_df, cor.y.abs <= glb_feats_df[glb_feats_df$id == ".rnorm", "cor.y.abs"])
# corxx_mtrx <- cor(data.matrix(glb_allobs_df[, setdiff(names(glb_allobs_df), myfind_chr_cols_df(glb_allobs_df))]), use="pairwise.complete.obs"); abs_corxx_mtrx <- abs(corxx_mtrx); diag(abs_corxx_mtrx) <- 0
# which.max(abs_corxx_mtrx["S.T.tribun", ])
# abs_corxx_mtrx["A.npnct08.log", "S.npnct08.log"]
# step_glm <- step(orig_glm)
# }
# Since caret does not optimize rpart well
# if (method == "rpart")
# ret_lst <- myfit_mdl(model_id=paste0(model_id_pfx, ".cp.0"), model_method=method,
# indep_vars_vctr=indep_vars_vctr,
# model_type=glb_model_type,
# rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
# fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
# n_cv_folds=0, tune_models_df=data.frame(parameter="cp", min=0.0, max=0.0, by=0.1))
}## label step_major step_minor bgn end elapsed
## 1 fit.models_1_bgn 1 0 190.320 190.336 0.016
## 2 fit.models_1_rpart 2 0 190.336 NA NA
## [1] "fitting model: All.X.no.rnorm.rpart"
## [1] " indep_vars: Age, Sex.fctr, Married.my.fctr, Region.fctr, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr"
## Aggregating results
## Selecting tuning parameters
## Fitting cp = 0.0193 on full training set## Warning in myfit_mdl(model_id = model_id, model_method = method,
## indep_vars_vctr = indep_vars_vctr, : model's bestTune found at an extreme
## of tuneGrid for parameter: cp
## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 0.25205411 0 1.0000000
## 2 0.09075572 1 0.7479459
## 3 0.01930707 2 0.6571902
##
## Variable importance
## Age Married.my.fctrWidowed
## 88 12
##
## Node number 1: 77145 observations, complexity param=0.2520541
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (62433 obs) right son=3 (14712 obs)
## Primary splits:
## Age < 64.5 to the left, improve=8861.839, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=2323.600, (0 missing)
## Married.my.fctrWidowed < 0.5 to the left, improve=2200.537, (0 missing)
## PeopleInHousehold < 2.5 to the right, improve=1947.586, (0 missing)
## Married.my.fctrNever Married < 0.5 to the right, improve=1420.837, (0 missing)
## Surrogate splits:
## Married.my.fctrWidowed < 0.5 to the left, agree=0.843, adj=0.178, (0 split)
##
## Node number 2: 62433 observations, complexity param=0.09075572
## predicted class=Employed expected loss=0.3215447 P(node) =0.8092942
## class counts: 3550 42358 10819 2767 2939
## probabilities: 0.057 0.678 0.173 0.044 0.047
## left son=4 (57429 obs) right son=5 (5004 obs)
## Primary splits:
## Age < 18.5 to the right, improve=3249.3410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1290.5850, (0 missing)
## Married.my.fctrNever Married < 0.5 to the left, improve=1036.8480, (0 missing)
## Married.my.fctrMarried < 0.5 to the right, improve= 695.1223, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 624.4119, (0 missing)
##
## Node number 3: 14712 observations
## predicted class=Retired expected loss=0.2627787 P(node) =0.1907058
## class counts: 626 2778 328 10846 134
## probabilities: 0.043 0.189 0.022 0.737 0.009
##
## Node number 4: 57429 observations
## predicted class=Employed expected loss=0.2781347 P(node) =0.7444293
## class counts: 3490 41456 7012 2765 2706
## probabilities: 0.061 0.722 0.122 0.048 0.047
##
## Node number 5: 5004 observations
## predicted class=Not.in.Labor.Force expected loss=0.2392086 P(node) =0.06486486
## class counts: 60 902 3807 2 233
## probabilities: 0.012 0.180 0.761 0.000 0.047
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 64.5 62433 20075 Employed (0.057 0.68 0.17 0.044 0.047)
## 4) Age>=18.5 57429 15973 Employed (0.061 0.72 0.12 0.048 0.047) *
## 5) Age< 18.5 5004 1197 Not.in.Labor.Force (0.012 0.18 0.76 0.0004 0.047) *
## 3) Age>=64.5 14712 3866 Retired (0.043 0.19 0.022 0.74 0.0091) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 3490 60 626
## Employed 0 41456 902 2778
## Not.in.Labor.Force 0 7012 3807 328
## Retired 0 2765 2 10846
## Unemployed 0 2706 233 134
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7273187 NA 0.7241611 0.7304591 0.5850800
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## [1] " calling mypredict_mdl for OOB:"
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 0 1289 21 226
## Employed 0 15206 365 1026
## Not.in.Labor.Force 0 2581 1416 102
## Retired 0 1028 1 3977
## Unemployed 0 986 97 47
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 0
## Retired 0
## Unemployed 0
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.7261351 0.4752624 0.7209052 0.7313187 0.5850606
## AccuracyPValue McnemarPValue
## 0.0000000 NaN
## model_id model_method
## 1 All.X.no.rnorm.rpart rpart
## feats
## 1 Age, Sex.fctr, Married.my.fctr, Region.fctr, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## max.nTuningRuns min.elapsedtime.everything min.elapsedtime.final
## 1 3 50.515 9.901
## max.Accuracy.fit max.AccuracyLower.fit max.AccuracyUpper.fit
## 1 0.73205 0.7241611 0.7304591
## max.Kappa.fit max.Accuracy.OOB max.AccuracyLower.OOB
## 1 0.4874172 0.7261351 0.7209052
## max.AccuracyUpper.OOB max.Kappa.OOB max.AccuracySD.fit max.KappaSD.fit
## 1 0.7313187 0.4752624 0.004250054 0.002519487# User specified
# easier to exclude features
#model_id_pfx <- "";
# indep_vars_vctr <- setdiff(names(glb_fitobs_df),
# union(union(glb_rsp_var, glb_exclude_vars_as_features),
# c("<feat1_name>", "<feat2_name>")))
# method <- ""
# easier to include features
# sav_models_lst <- glb_models_lst; sav_models_df <- glb_models_df; sav_featsimp_df <- glb_featsimp_df
# glb_models_lst <- sav_models_lst; glb_models_df <- sav_models_df; glm_featsimp_df <- sav_featsimp_df
#table(glb_allobs_df$myCategory, glb_allobs_df$H.P.readers.respond, glb_allobs_df[, glb_rsp_var], useNA="ifany")
#model_id <- "Rank9.2"; indep_vars_vctr <- c(NULL
# ,"<feat1>"
# ,"<feat1>*<feat2>"
# ,"<feat1>:<feat2>"
# )
# for (method in c("bayesglm")) {
# ret_lst <- myfit_mdl(model_id=model_id, model_method=method,
# indep_vars_vctr=indep_vars_vctr,
# model_type=glb_model_type,
# rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
# fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
# n_cv_folds=glb_n_cv_folds, tune_models_df=glb_tune_models_df)
# # csm_mdl_id <- paste0(model_id, ".", method)
# # csm_featsimp_df <- myget_feats_importance(glb_models_lst[[paste0(model_id, ".", method)]]); print(head(csm_featsimp_df))
# }
#csm_featsimp_df[grepl("H.npnct19.log", row.names(csm_featsimp_df)), , FALSE]
#csm_OOBobs_df <- glb_get_predictions(glb_OOBobs_df, mdl_id=csm_mdl_id, rsp_var_out=glb_rsp_var_out, prob_threshold_def=glb_models_df[glb_models_df$model_id == csm_mdl_id, "opt.prob.threshold.OOB"])
#print(sprintf("%s OOB confusion matrix & accuracy: ", csm_mdl_id)); print(t(confusionMatrix(csm_OOBobs_df[, paste0(glb_rsp_var_out, csm_mdl_id)], csm_OOBobs_df[, glb_rsp_var])$table))
#glb_models_df[, "max.Accuracy.OOB", FALSE]
#varImp(glb_models_lst[["Low.cor.X.glm"]])
#orderBy(~ -Overall, varImp(glb_models_lst[["All.X.2.glm"]])$importance)
#orderBy(~ -Overall, varImp(glb_models_lst[["All.X.3.glm"]])$importance)
#glb_feats_df[grepl("npnct28", glb_feats_df$id), ]
#print(sprintf("%s OOB confusion matrix & accuracy: ", glb_sel_mdl_id)); print(t(confusionMatrix(glb_OOBobs_df[, paste0(glb_rsp_var_out, glb_sel_mdl_id)], glb_OOBobs_df[, glb_rsp_var])$table))
# User specified bivariate models
# indep_vars_vctr_lst <- list()
# for (feat in setdiff(names(glb_fitobs_df),
# union(glb_rsp_var, glb_exclude_vars_as_features)))
# indep_vars_vctr_lst[["feat"]] <- feat
# User specified combinatorial models
# indep_vars_vctr_lst <- list()
# combn_mtrx <- combn(c("<feat1_name>", "<feat2_name>", "<featn_name>"),
# <num_feats_to_choose>)
# for (combn_ix in 1:ncol(combn_mtrx))
# #print(combn_mtrx[, combn_ix])
# indep_vars_vctr_lst[[combn_ix]] <- combn_mtrx[, combn_ix]
# template for myfit_mdl
# rf is hard-coded in caret to recognize only Accuracy / Kappa evaluation metrics
# only for OOB in trainControl ?
# ret_lst <- myfit_mdl_fn(model_id=paste0(model_id_pfx, ""), model_method=method,
# indep_vars_vctr=indep_vars_vctr,
# rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
# fit_df=glb_fitobs_df, OOB_df=glb_OOBobs_df,
# n_cv_folds=glb_n_cv_folds, tune_models_df=glb_tune_models_df,
# model_loss_mtrx=glb_model_metric_terms,
# model_summaryFunction=glb_model_metric_smmry,
# model_metric=glb_model_metric,
# model_metric_maximize=glb_model_metric_maximize)
# Simplify a model
# fit_df <- glb_fitobs_df; glb_mdl <- step(<complex>_mdl)
# Non-caret models
# rpart_area_mdl <- rpart(reformulate("Area", response=glb_rsp_var),
# data=glb_fitobs_df, #method="class",
# control=rpart.control(cp=0.12),
# parms=list(loss=glb_model_metric_terms))
# print("rpart_sel_wlm_mdl"); prp(rpart_sel_wlm_mdl)
#
print(glb_models_df)## model_id model_method
## MFO.myMFO_classfr MFO.myMFO_classfr myMFO_classfr
## Random.myrandom_classfr Random.myrandom_classfr myrandom_classfr
## Max.cor.Y.cv.0.rpart Max.cor.Y.cv.0.rpart rpart
## Max.cor.Y.cv.0.cp.0.rpart Max.cor.Y.cv.0.cp.0.rpart rpart
## Max.cor.Y.rpart Max.cor.Y.rpart rpart
## Low.cor.X.rpart Low.cor.X.rpart rpart
## All.X.no.rnorm.rpart All.X.no.rnorm.rpart rpart
## feats
## MFO.myMFO_classfr .rnorm
## Random.myrandom_classfr .rnorm
## Max.cor.Y.cv.0.rpart Industry.my.fctr, Age
## Max.cor.Y.cv.0.cp.0.rpart Industry.my.fctr, Age
## Max.cor.Y.rpart Industry.my.fctr, Age
## Low.cor.X.rpart Age, Sex.fctr, Married.my.fctr, Region.fctr, .rnorm, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## All.X.no.rnorm.rpart Age, Sex.fctr, Married.my.fctr, Region.fctr, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## max.nTuningRuns min.elapsedtime.everything
## MFO.myMFO_classfr 0 0.471
## Random.myrandom_classfr 0 0.304
## Max.cor.Y.cv.0.rpart 0 13.775
## Max.cor.Y.cv.0.cp.0.rpart 0 4.443
## Max.cor.Y.rpart 3 17.853
## Low.cor.X.rpart 3 57.322
## All.X.no.rnorm.rpart 3 50.515
## min.elapsedtime.final max.Accuracy.fit
## MFO.myMFO_classfr 0.012 0.5850800
## Random.myrandom_classfr 0.010 0.3982241
## Max.cor.Y.cv.0.rpart 4.064 0.5850800
## Max.cor.Y.cv.0.cp.0.rpart 3.816 0.8832199
## Max.cor.Y.rpart 3.830 0.7320500
## Low.cor.X.rpart 10.976 0.7320500
## All.X.no.rnorm.rpart 9.901 0.7320500
## max.AccuracyLower.fit max.AccuracyUpper.fit
## MFO.myMFO_classfr 0.5815936 0.5885601
## Random.myrandom_classfr 0.3947671 0.4016888
## Max.cor.Y.cv.0.rpart 0.5815936 0.5885601
## Max.cor.Y.cv.0.cp.0.rpart 0.8809327 0.8854782
## Max.cor.Y.rpart 0.7241611 0.7304591
## Low.cor.X.rpart 0.7241611 0.7304591
## All.X.no.rnorm.rpart 0.7241611 0.7304591
## max.Kappa.fit max.Accuracy.OOB
## MFO.myMFO_classfr NA 0.5850606
## Random.myrandom_classfr NA 0.3982657
## Max.cor.Y.cv.0.rpart NA 0.5850606
## Max.cor.Y.cv.0.cp.0.rpart NA 0.8828257
## Max.cor.Y.rpart 0.4874172 0.7261351
## Low.cor.X.rpart 0.4874172 0.7261351
## All.X.no.rnorm.rpart 0.4874172 0.7261351
## max.AccuracyLower.OOB max.AccuracyUpper.OOB
## MFO.myMFO_classfr 0.5793009 0.5908029
## Random.myrandom_classfr 0.3925620 0.4039901
## Max.cor.Y.cv.0.rpart 0.5793009 0.5908029
## Max.cor.Y.cv.0.cp.0.rpart 0.8790261 0.8865467
## Max.cor.Y.rpart 0.7209052 0.7313187
## Low.cor.X.rpart 0.7209052 0.7313187
## All.X.no.rnorm.rpart 0.7209052 0.7313187
## max.Kappa.OOB max.AccuracySD.fit max.KappaSD.fit
## MFO.myMFO_classfr 0.0000000000 NA NA
## Random.myrandom_classfr -0.0008038519 NA NA
## Max.cor.Y.cv.0.rpart 0.0000000000 NA NA
## Max.cor.Y.cv.0.cp.0.rpart 0.7967286427 NA NA
## Max.cor.Y.rpart 0.4752624489 0.004250054 0.002519487
## Low.cor.X.rpart 0.4752624489 0.004250054 0.002519487
## All.X.no.rnorm.rpart 0.4752624489 0.004250054 0.002519487rm(ret_lst)
fit.models_1_chunk_df <- myadd_chunk(fit.models_1_chunk_df, "fit.models_1_end",
major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 2 fit.models_1_rpart 2 0 190.336 244.878 54.542
## 3 fit.models_1_end 3 0 244.879 NA NAglb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.models", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 11 fit.models 7 1 186.412 244.918 58.506
## 12 fit.models 7 2 244.919 NA NAif (!is.null(glb_model_metric_smmry)) {
stats_df <- glb_models_df[, "model_id", FALSE]
stats_mdl_df <- data.frame()
for (model_id in stats_df$model_id) {
stats_mdl_df <- rbind(stats_mdl_df,
mypredict_mdl(glb_models_lst[[model_id]], glb_fitobs_df, glb_rsp_var,
glb_rsp_var_out, model_id, "fit",
glb_model_metric_smmry, glb_model_metric,
glb_model_metric_maximize, ret_type="stats"))
}
stats_df <- merge(stats_df, stats_mdl_df, all.x=TRUE)
stats_mdl_df <- data.frame()
for (model_id in stats_df$model_id) {
stats_mdl_df <- rbind(stats_mdl_df,
mypredict_mdl(glb_models_lst[[model_id]], glb_OOBobs_df, glb_rsp_var,
glb_rsp_var_out, model_id, "OOB",
glb_model_metric_smmry, glb_model_metric,
glb_model_metric_maximize, ret_type="stats"))
}
stats_df <- merge(stats_df, stats_mdl_df, all.x=TRUE)
# tmp_models_df <- orderBy(~model_id, glb_models_df)
# rownames(tmp_models_df) <- seq(1, nrow(tmp_models_df))
# all.equal(subset(tmp_models_df[, names(stats_df)], model_id != "Random.myrandom_classfr"),
# subset(stats_df, model_id != "Random.myrandom_classfr"))
# print(subset(tmp_models_df[, names(stats_df)], model_id != "Random.myrandom_classfr")[, c("model_id", "max.Accuracy.fit")])
# print(subset(stats_df, model_id != "Random.myrandom_classfr")[, c("model_id", "max.Accuracy.fit")])
print("Merging following data into glb_models_df:")
print(stats_mrg_df <- stats_df[, c(1, grep(glb_model_metric, names(stats_df)))])
print(tmp_models_df <- orderBy(~model_id, glb_models_df[, c("model_id", grep(glb_model_metric, names(stats_df), value=TRUE))]))
tmp2_models_df <- glb_models_df[, c("model_id", setdiff(names(glb_models_df), grep(glb_model_metric, names(stats_df), value=TRUE)))]
tmp3_models_df <- merge(tmp2_models_df, stats_mrg_df, all.x=TRUE, sort=FALSE)
print(tmp3_models_df)
print(names(tmp3_models_df))
print(glb_models_df <- subset(tmp3_models_df, select=-model_id.1))
}
plt_models_df <- glb_models_df[, -grep("SD|Upper|Lower", names(glb_models_df))]
for (var in grep("^min.", names(plt_models_df), value=TRUE)) {
plt_models_df[, sub("min.", "inv.", var)] <-
#ifelse(all(is.na(tmp <- plt_models_df[, var])), NA, 1.0 / tmp)
1.0 / plt_models_df[, var]
plt_models_df <- plt_models_df[ , -grep(var, names(plt_models_df))]
}
print(plt_models_df)## model_id model_method
## MFO.myMFO_classfr MFO.myMFO_classfr myMFO_classfr
## Random.myrandom_classfr Random.myrandom_classfr myrandom_classfr
## Max.cor.Y.cv.0.rpart Max.cor.Y.cv.0.rpart rpart
## Max.cor.Y.cv.0.cp.0.rpart Max.cor.Y.cv.0.cp.0.rpart rpart
## Max.cor.Y.rpart Max.cor.Y.rpart rpart
## Low.cor.X.rpart Low.cor.X.rpart rpart
## All.X.no.rnorm.rpart All.X.no.rnorm.rpart rpart
## feats
## MFO.myMFO_classfr .rnorm
## Random.myrandom_classfr .rnorm
## Max.cor.Y.cv.0.rpart Industry.my.fctr, Age
## Max.cor.Y.cv.0.cp.0.rpart Industry.my.fctr, Age
## Max.cor.Y.rpart Industry.my.fctr, Age
## Low.cor.X.rpart Age, Sex.fctr, Married.my.fctr, Region.fctr, .rnorm, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## All.X.no.rnorm.rpart Age, Sex.fctr, Married.my.fctr, Region.fctr, Race.fctr, Citizenship.fctr, Hispanic, PeopleInHousehold, Education.my.fctr, Industry.my.fctr
## max.nTuningRuns max.Accuracy.fit max.Kappa.fit
## MFO.myMFO_classfr 0 0.5850800 NA
## Random.myrandom_classfr 0 0.3982241 NA
## Max.cor.Y.cv.0.rpart 0 0.5850800 NA
## Max.cor.Y.cv.0.cp.0.rpart 0 0.8832199 NA
## Max.cor.Y.rpart 3 0.7320500 0.4874172
## Low.cor.X.rpart 3 0.7320500 0.4874172
## All.X.no.rnorm.rpart 3 0.7320500 0.4874172
## max.Accuracy.OOB max.Kappa.OOB
## MFO.myMFO_classfr 0.5850606 0.0000000000
## Random.myrandom_classfr 0.3982657 -0.0008038519
## Max.cor.Y.cv.0.rpart 0.5850606 0.0000000000
## Max.cor.Y.cv.0.cp.0.rpart 0.8828257 0.7967286427
## Max.cor.Y.rpart 0.7261351 0.4752624489
## Low.cor.X.rpart 0.7261351 0.4752624489
## All.X.no.rnorm.rpart 0.7261351 0.4752624489
## inv.elapsedtime.everything inv.elapsedtime.final
## MFO.myMFO_classfr 2.12314225 83.33333333
## Random.myrandom_classfr 3.28947368 100.00000000
## Max.cor.Y.cv.0.rpart 0.07259528 0.24606299
## Max.cor.Y.cv.0.cp.0.rpart 0.22507315 0.26205451
## Max.cor.Y.rpart 0.05601300 0.26109661
## Low.cor.X.rpart 0.01744531 0.09110787
## All.X.no.rnorm.rpart 0.01979610 0.10099990print(myplot_radar(radar_inp_df=plt_models_df))## Warning: The shape palette can deal with a maximum of 6 discrete values
## because more than 6 becomes difficult to discriminate; you have 7.
## Consider specifying shapes manually. if you must have them.## Warning in loop_apply(n, do.ply): Removed 11 rows containing missing values
## (geom_point).## Warning in loop_apply(n, do.ply): Removed 4 rows containing missing values
## (geom_text).## Warning: The shape palette can deal with a maximum of 6 discrete values
## because more than 6 becomes difficult to discriminate; you have 7.
## Consider specifying shapes manually. if you must have them.
# print(myplot_radar(radar_inp_df=subset(plt_models_df,
# !(model_id %in% grep("random|MFO", plt_models_df$model_id, value=TRUE)))))
# Compute CI for <metric>SD
glb_models_df <- mutate(glb_models_df,
max.df = ifelse(max.nTuningRuns > 1, max.nTuningRuns - 1, NA),
min.sd2ci.scaler = ifelse(is.na(max.df), NA, qt(0.975, max.df)))
for (var in grep("SD", names(glb_models_df), value=TRUE)) {
# Does CI alredy exist ?
var_components <- unlist(strsplit(var, "SD"))
varActul <- paste0(var_components[1], var_components[2])
varUpper <- paste0(var_components[1], "Upper", var_components[2])
varLower <- paste0(var_components[1], "Lower", var_components[2])
if (varUpper %in% names(glb_models_df)) {
warning(varUpper, " already exists in glb_models_df")
# Assuming Lower also exists
next
}
print(sprintf("var:%s", var))
# CI is dependent on sample size in t distribution; df=n-1
glb_models_df[, varUpper] <- glb_models_df[, varActul] +
glb_models_df[, "min.sd2ci.scaler"] * glb_models_df[, var]
glb_models_df[, varLower] <- glb_models_df[, varActul] -
glb_models_df[, "min.sd2ci.scaler"] * glb_models_df[, var]
}## Warning: max.AccuracyUpper.fit already exists in glb_models_df## [1] "var:max.KappaSD.fit"# Plot metrics with CI
plt_models_df <- glb_models_df[, "model_id", FALSE]
pltCI_models_df <- glb_models_df[, "model_id", FALSE]
for (var in grep("Upper", names(glb_models_df), value=TRUE)) {
var_components <- unlist(strsplit(var, "Upper"))
col_name <- unlist(paste(var_components, collapse=""))
plt_models_df[, col_name] <- glb_models_df[, col_name]
for (name in paste0(var_components[1], c("Upper", "Lower"), var_components[2]))
pltCI_models_df[, name] <- glb_models_df[, name]
}
build_statsCI_data <- function(plt_models_df) {
mltd_models_df <- melt(plt_models_df, id.vars="model_id")
mltd_models_df$data <- sapply(1:nrow(mltd_models_df),
function(row_ix) tail(unlist(strsplit(as.character(
mltd_models_df[row_ix, "variable"]), "[.]")), 1))
mltd_models_df$label <- sapply(1:nrow(mltd_models_df),
function(row_ix) head(unlist(strsplit(as.character(
mltd_models_df[row_ix, "variable"]),
paste0(".", mltd_models_df[row_ix, "data"]))), 1))
#print(mltd_models_df)
return(mltd_models_df)
}
mltd_models_df <- build_statsCI_data(plt_models_df)
mltdCI_models_df <- melt(pltCI_models_df, id.vars="model_id")
for (row_ix in 1:nrow(mltdCI_models_df)) {
for (type in c("Upper", "Lower")) {
if (length(var_components <- unlist(strsplit(
as.character(mltdCI_models_df[row_ix, "variable"]), type))) > 1) {
#print(sprintf("row_ix:%d; type:%s; ", row_ix, type))
mltdCI_models_df[row_ix, "label"] <- var_components[1]
mltdCI_models_df[row_ix, "data"] <-
unlist(strsplit(var_components[2], "[.]"))[2]
mltdCI_models_df[row_ix, "type"] <- type
break
}
}
}
#print(mltdCI_models_df)
# castCI_models_df <- dcast(mltdCI_models_df, value ~ type, fun.aggregate=sum)
# print(castCI_models_df)
wideCI_models_df <- reshape(subset(mltdCI_models_df, select=-variable),
timevar="type",
idvar=setdiff(names(mltdCI_models_df), c("type", "value", "variable")),
direction="wide")
#print(wideCI_models_df)
mrgdCI_models_df <- merge(wideCI_models_df, mltd_models_df, all.x=TRUE)
#print(mrgdCI_models_df)
# Merge stats back in if CIs don't exist
goback_vars <- c()
for (var in unique(mltd_models_df$label)) {
for (type in unique(mltd_models_df$data)) {
var_type <- paste0(var, ".", type)
# if this data is already present, next
if (var_type %in% unique(paste(mltd_models_df$label, mltd_models_df$data,
sep=".")))
next
#print(sprintf("var_type:%s", var_type))
goback_vars <- c(goback_vars, var_type)
}
}
if (length(goback_vars) > 0) {
mltd_goback_df <- build_statsCI_data(glb_models_df[, c("model_id", goback_vars)])
mltd_models_df <- rbind(mltd_models_df, mltd_goback_df)
}
mltd_models_df <- merge(mltd_models_df, glb_models_df[, c("model_id", "model_method")],
all.x=TRUE)
png(paste0(glb_out_pfx, "models_bar.png"), width=480*3, height=480*2)
print(gp <- myplot_bar(mltd_models_df, "model_id", "value", colorcol_name="model_method") +
geom_errorbar(data=mrgdCI_models_df,
mapping=aes(x=model_id, ymax=value.Upper, ymin=value.Lower), width=0.5) +
facet_grid(label ~ data, scales="free") +
theme(axis.text.x = element_text(angle = 90,vjust = 0.5)))## Warning in loop_apply(n, do.ply): Removed 4 rows containing missing values
## (position_stack).## Warning in loop_apply(n, do.ply): Stacking not well defined when ymin != 0dev.off()## quartz_off_screen
## 2print(gp)## Warning in loop_apply(n, do.ply): Removed 4 rows containing missing values
## (position_stack).## Warning in loop_apply(n, do.ply): Stacking not well defined when ymin != 0
# used for console inspection
model_evl_terms <- c(NULL)
for (metric in glb_model_evl_criteria)
model_evl_terms <- c(model_evl_terms,
ifelse(length(grep("max", metric)) > 0, "-", "+"), metric)
if (glb_is_classification && glb_is_binomial)
model_evl_terms <- c(model_evl_terms, "-", "opt.prob.threshold.OOB")
model_sel_frmla <- as.formula(paste(c("~ ", model_evl_terms), collapse=" "))
dsp_models_cols <- c("model_id", glb_model_evl_criteria)
if (glb_is_classification && glb_is_binomial)
dsp_models_cols <- c(dsp_models_cols, "opt.prob.threshold.OOB")
print(dsp_models_df <- orderBy(model_sel_frmla, glb_models_df)[, dsp_models_cols])## model_id max.Accuracy.OOB max.Kappa.OOB
## 4 Max.cor.Y.cv.0.cp.0.rpart 0.8828257 0.7967286427
## 5 Max.cor.Y.rpart 0.7261351 0.4752624489
## 6 Low.cor.X.rpart 0.7261351 0.4752624489
## 7 All.X.no.rnorm.rpart 0.7261351 0.4752624489
## 1 MFO.myMFO_classfr 0.5850606 0.0000000000
## 3 Max.cor.Y.cv.0.rpart 0.5850606 0.0000000000
## 2 Random.myrandom_classfr 0.3982657 -0.0008038519print(myplot_radar(radar_inp_df=dsp_models_df))## Warning: The shape palette can deal with a maximum of 6 discrete values
## because more than 6 becomes difficult to discriminate; you have 7.
## Consider specifying shapes manually. if you must have them.## Warning in loop_apply(n, do.ply): Removed 3 rows containing missing values
## (geom_point).## Warning: The shape palette can deal with a maximum of 6 discrete values
## because more than 6 becomes difficult to discriminate; you have 7.
## Consider specifying shapes manually. if you must have them.
print("Metrics used for model selection:"); print(model_sel_frmla)## [1] "Metrics used for model selection:"## ~-max.Accuracy.OOB - max.Kappa.OOBprint(sprintf("Best model id: %s", dsp_models_df[1, "model_id"]))## [1] "Best model id: Max.cor.Y.cv.0.cp.0.rpart"if (is.null(glb_sel_mdl_id)) {
glb_sel_mdl_id <- dsp_models_df[1, "model_id"]
if (glb_sel_mdl_id == "Interact.High.cor.Y.glm") {
warning("glb_sel_mdl_id: Interact.High.cor.Y.glm; myextract_mdl_feats does not currently support interaction terms")
glb_sel_mdl_id <- dsp_models_df[2, "model_id"]
}
} else
print(sprintf("User specified selection: %s", glb_sel_mdl_id))
myprint_mdl(glb_sel_mdl <- glb_models_lst[[glb_sel_mdl_id]])
## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 77145
##
## CP nsplit rel error
## 1 2.520541e-01 0 1.0000000
## 2 9.075572e-02 1 0.7479459
## 3 1.930707e-02 2 0.6571902
## 4 1.190290e-02 3 0.6378831
## 5 1.134056e-02 4 0.6259802
## 6 1.065325e-02 5 0.6146396
## 7 9.256772e-03 6 0.6039864
## 8 7.591615e-03 19 0.4122903
## 9 6.310725e-03 20 0.4046987
## 10 6.248243e-03 22 0.3920772
## 11 6.217001e-03 23 0.3858290
## 12 6.092037e-03 24 0.3796120
## 13 6.029554e-03 25 0.3735199
## 14 4.717423e-03 26 0.3674904
## 15 4.592458e-03 27 0.3627730
## 16 4.092599e-03 28 0.3581805
## 17 3.842669e-03 29 0.3540879
## 18 3.811428e-03 30 0.3502452
## 19 3.249086e-03 31 0.3464338
## 20 3.238672e-03 32 0.3431847
## 21 2.936674e-03 41 0.2993221
## 22 2.530538e-03 42 0.2963854
## 23 1.749508e-03 43 0.2938549
## 24 1.687026e-03 44 0.2921053
## 25 1.499578e-03 45 0.2904183
## 26 1.030960e-03 46 0.2889187
## 27 9.684776e-04 47 0.2878878
## 28 9.059952e-04 49 0.2859508
## 29 7.497891e-04 50 0.2850448
## 30 6.248243e-04 51 0.2842950
## 31 4.998594e-04 52 0.2836702
## 32 4.373770e-04 53 0.2831704
## 33 3.748946e-04 54 0.2827330
## 34 2.030679e-04 55 0.2823581
## 35 1.718267e-04 57 0.2819520
## 36 6.248243e-05 59 0.2816083
## 37 3.124121e-05 60 0.2815458
## 38 0.000000e+00 63 0.2814521
##
## Variable importance
## Age
## 43
## Industry.my.fctrOther services
## 7
## Industry.my.fctrPublic administration
## 5
## Industry.my.fctrEducational and health services
## 5
## Industry.my.fctrLeisure and hospitality
## 5
## Industry.my.fctrTransportation and utilities
## 5
## Industry.my.fctrTrade
## 5
## Industry.my.fctrConstruction
## 4
## Industry.my.fctrFinancial
## 4
## Industry.my.fctrProfessional and business services
## 4
## Industry.my.fctrManufacturing
## 4
## Industry.my.fctrAgriculture, forestry, fishing, and hunting
## 4
## Industry.my.fctrInformation
## 3
## Industry.my.fctrMining
## 2
##
## Node number 1: 77145 observations, complexity param=0.2520541
## predicted class=Employed expected loss=0.41492 P(node) =1
## class counts: 4176 45136 11147 13613 3073
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (62433 obs) right son=3 (14712 obs)
## Primary splits:
## Age < 64.5 to the left, improve=8861.8390, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=2323.6000, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=1186.7910, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 983.1942, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 924.0788, (0 missing)
##
## Node number 2: 62433 observations, complexity param=0.09075572
## predicted class=Employed expected loss=0.3215447 P(node) =0.8092942
## class counts: 3550 42358 10819 2767 2939
## probabilities: 0.057 0.678 0.173 0.044 0.047
## left son=4 (57429 obs) right son=5 (5004 obs)
## Primary splits:
## Age < 18.5 to the right, improve=3249.3410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1290.5850, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 624.4119, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 515.9783, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 505.1749, (0 missing)
##
## Node number 3: 14712 observations, complexity param=0.01930707
## predicted class=Retired expected loss=0.2627787 P(node) =0.1907058
## class counts: 626 2778 328 10846 134
## probabilities: 0.043 0.189 0.022 0.737 0.009
## left son=6 (696 obs) right son=7 (14016 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=756.4649, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=436.5917, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=408.2808, (0 missing)
## Age < 70.5 to the left, improve=364.4106, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=241.2709, (0 missing)
##
## Node number 4: 57429 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.2781347 P(node) =0.7444293
## class counts: 3490 41456 7012 2765 2706
## probabilities: 0.061 0.722 0.122 0.048 0.047
## left son=8 (10137 obs) right son=9 (47292 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=880.1351, (0 missing)
## Age < 59.5 to the left, improve=604.0624, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=395.1759, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=341.2702, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=334.6625, (0 missing)
##
## Node number 5: 5004 observations, complexity param=0.0119029
## predicted class=Not.in.Labor.Force expected loss=0.2392086 P(node) =0.06486486
## class counts: 60 902 3807 2 233
## probabilities: 0.012 0.180 0.761 0.000 0.047
## left son=10 (487 obs) right son=11 (4517 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=501.58970, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=243.57050, (0 missing)
## Age < 16.5 to the right, improve=107.27410, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve= 66.39350, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 38.80304, (0 missing)
##
## Node number 6: 696 observations
## predicted class=Employed expected loss=0.07758621 P(node) =0.009021972
## class counts: 0 642 0 24 30
## probabilities: 0.000 0.922 0.000 0.034 0.043
##
## Node number 7: 14016 observations, complexity param=0.01134056
## predicted class=Retired expected loss=0.2278824 P(node) =0.1816838
## class counts: 626 2136 328 10822 104
## probabilities: 0.045 0.152 0.023 0.772 0.007
## left son=14 (402 obs) right son=15 (13614 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=481.2436, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=450.4300, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=265.4408, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=262.8831, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=260.1050, (0 missing)
##
## Node number 8: 10137 observations
## predicted class=Employed expected loss=0.05011345 P(node) =0.1314019
## class counts: 9 9629 51 13 435
## probabilities: 0.001 0.950 0.005 0.001 0.043
##
## Node number 9: 47292 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3270109 P(node) =0.6130274
## class counts: 3481 31827 6961 2752 2271
## probabilities: 0.074 0.673 0.147 0.058 0.048
## left son=18 (42249 obs) right son=19 (5043 obs)
## Primary splits:
## Age < 59.5 to the left, improve=682.6394, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=615.9417, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=526.5017, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=511.4015, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=354.9892, (0 missing)
##
## Node number 10: 487 observations
## predicted class=Employed expected loss=0.1457906 P(node) =0.006312788
## class counts: 0 416 35 0 36
## probabilities: 0.000 0.854 0.072 0.000 0.074
##
## Node number 11: 4517 observations, complexity param=0.006029554
## predicted class=Not.in.Labor.Force expected loss=0.1649325 P(node) =0.05855208
## class counts: 60 486 3772 2 197
## probabilities: 0.013 0.108 0.835 0.000 0.044
## left son=22 (232 obs) right son=23 (4285 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=298.35260, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve= 84.49187, (0 missing)
## Age < 16.5 to the right, improve= 55.14438, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 49.12110, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 48.50088, (0 missing)
##
## Node number 14: 402 observations
## predicted class=Employed expected loss=0.06716418 P(node) =0.005210966
## class counts: 2 375 0 12 13
## probabilities: 0.005 0.933 0.000 0.030 0.032
##
## Node number 15: 13614 observations, complexity param=0.01065325
## predicted class=Retired expected loss=0.2059644 P(node) =0.1764729
## class counts: 624 1761 328 10810 91
## probabilities: 0.046 0.129 0.024 0.794 0.007
## left son=30 (384 obs) right son=31 (13230 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=478.0754, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=281.2735, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=278.5598, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=275.8898, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=270.8294, (0 missing)
##
## Node number 18: 42249 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3028001 P(node) =0.547657
## class counts: 2844 29456 6676 1140 2133
## probabilities: 0.067 0.697 0.158 0.027 0.050
## left son=36 (5469 obs) right son=37 (36780 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=491.7069, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=427.5658, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=419.4708, (0 missing)
## Age < 23.5 to the right, improve=324.6411, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=289.1954, (0 missing)
##
## Node number 19: 5043 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.5298433 P(node) =0.06537041
## class counts: 637 2371 285 1612 138
## probabilities: 0.126 0.470 0.057 0.320 0.027
## left son=38 (451 obs) right son=39 (4592 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=169.22760, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=123.99940, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=118.31870, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 90.86328, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 76.67119, (0 missing)
##
## Node number 22: 232 observations
## predicted class=Employed expected loss=0.1336207 P(node) =0.003007324
## class counts: 0 201 8 0 23
## probabilities: 0.000 0.866 0.034 0.000 0.099
##
## Node number 23: 4285 observations, complexity param=0.001749508
## predicted class=Not.in.Labor.Force expected loss=0.1215869 P(node) =0.05554475
## class counts: 60 285 3764 2 174
## probabilities: 0.014 0.067 0.878 0.000 0.041
## left son=46 (97 obs) right son=47 (4188 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=95.90269, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=55.61024, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=55.21431, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=53.14430, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=39.60641, (0 missing)
##
## Node number 30: 384 observations
## predicted class=Employed expected loss=0.0859375 P(node) =0.00497764
## class counts: 1 351 2 10 20
## probabilities: 0.003 0.914 0.005 0.026 0.052
##
## Node number 31: 13230 observations, complexity param=0.006310725
## predicted class=Retired expected loss=0.1836735 P(node) =0.1714952
## class counts: 623 1410 326 10800 71
## probabilities: 0.047 0.107 0.025 0.816 0.005
## left son=62 (219 obs) right son=63 (13011 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=297.6126, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=294.7483, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=292.1900, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=286.7155, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=222.5232, (0 missing)
##
## Node number 36: 5469 observations
## predicted class=Employed expected loss=0.07899067 P(node) =0.07089248
## class counts: 7 5037 40 1 384
## probabilities: 0.001 0.921 0.007 0.000 0.070
##
## Node number 37: 36780 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3360794 P(node) =0.4767645
## class counts: 2837 24419 6636 1139 1749
## probabilities: 0.077 0.664 0.180 0.031 0.048
## left son=74 (4668 obs) right son=75 (32112 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=571.8024, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=556.1714, (0 missing)
## Age < 23.5 to the right, improve=394.3466, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=392.5422, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=371.2783, (0 missing)
##
## Node number 38: 451 observations
## predicted class=Employed expected loss=0.05543237 P(node) =0.005846134
## class counts: 1 426 1 3 20
## probabilities: 0.002 0.945 0.002 0.007 0.044
##
## Node number 39: 4592 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.5764373 P(node) =0.05952427
## class counts: 636 1945 284 1609 118
## probabilities: 0.139 0.424 0.062 0.350 0.026
## left son=78 (354 obs) right son=79 (4238 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=151.30200, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=145.26730, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=109.69160, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 93.33870, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve= 76.52206, (0 missing)
##
## Node number 46: 97 observations
## predicted class=Employed expected loss=0.2474227 P(node) =0.001257372
## class counts: 0 73 17 0 7
## probabilities: 0.000 0.753 0.175 0.000 0.072
##
## Node number 47: 4188 observations, complexity param=0.00103096
## predicted class=Not.in.Labor.Force expected loss=0.1053009 P(node) =0.05428738
## class counts: 60 212 3747 2 167
## probabilities: 0.014 0.051 0.895 0.000 0.040
## left son=94 (55 obs) right son=95 (4133 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=58.19129, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=57.88160, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=55.33508, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=41.20103, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=37.29178, (0 missing)
##
## Node number 62: 219 observations
## predicted class=Employed expected loss=0.05479452 P(node) =0.00283881
## class counts: 0 207 0 5 7
## probabilities: 0.000 0.945 0.000 0.023 0.032
##
## Node number 63: 13011 observations, complexity param=0.006248243
## predicted class=Retired expected loss=0.1703174 P(node) =0.1686564
## class counts: 623 1203 326 10795 64
## probabilities: 0.048 0.092 0.025 0.830 0.005
## left son=126 (217 obs) right son=127 (12794 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=304.8379, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=302.3581, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=296.6215, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=229.9711, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=198.7685, (0 missing)
##
## Node number 74: 4668 observations
## predicted class=Employed expected loss=0.07219366 P(node) =0.06050943
## class counts: 3 4331 31 0 303
## probabilities: 0.001 0.928 0.007 0.000 0.065
##
## Node number 75: 32112 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.3744395 P(node) =0.4162551
## class counts: 2834 20088 6605 1139 1446
## probabilities: 0.088 0.626 0.206 0.035 0.045
## left son=150 (4364 obs) right son=151 (27748 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=739.5833, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=532.0889, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=481.7038, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=415.8067, (0 missing)
## Age < 23.5 to the right, improve=364.5501, (0 missing)
##
## Node number 78: 354 observations
## predicted class=Employed expected loss=0.07344633 P(node) =0.004588761
## class counts: 0 328 0 3 23
## probabilities: 0.000 0.927 0.000 0.008 0.065
##
## Node number 79: 4238 observations, complexity param=0.003238672
## predicted class=Employed expected loss=0.6184521 P(node) =0.05493551
## class counts: 636 1617 284 1606 95
## probabilities: 0.150 0.382 0.067 0.379 0.022
## left son=158 (361 obs) right son=159 (3877 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=172.59900, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=128.53290, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=110.07090, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve= 90.67191, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 88.92407, (0 missing)
##
## Node number 94: 55 observations
## predicted class=Employed expected loss=0.2363636 P(node) =0.0007129432
## class counts: 0 42 9 0 4
## probabilities: 0.000 0.764 0.164 0.000 0.073
##
## Node number 95: 4133 observations, complexity param=0.0009684776
## predicted class=Not.in.Labor.Force expected loss=0.09557222 P(node) =0.05357444
## class counts: 60 170 3738 2 163
## probabilities: 0.015 0.041 0.904 0.000 0.039
## left son=190 (61 obs) right son=191 (4072 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=59.50594, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=56.66486, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=42.16854, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=38.28705, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=27.80150, (0 missing)
##
## Node number 126: 217 observations
## predicted class=Employed expected loss=0.06912442 P(node) =0.002812885
## class counts: 0 202 1 2 12
## probabilities: 0.000 0.931 0.005 0.009 0.055
##
## Node number 127: 12794 observations, complexity param=0.006217001
## predicted class=Retired expected loss=0.1564014 P(node) =0.1658435
## class counts: 623 1001 325 10793 52
## probabilities: 0.049 0.078 0.025 0.844 0.004
## left son=254 (222 obs) right son=255 (12572 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=312.9565, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=306.9477, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=237.7246, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=205.6450, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=192.0610, (0 missing)
##
## Node number 150: 4364 observations
## predicted class=Employed expected loss=0.06393217 P(node) =0.0565688
## class counts: 2 4085 15 3 259
## probabilities: 0.000 0.936 0.003 0.001 0.059
##
## Node number 151: 27748 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.4232737 P(node) =0.3596863
## class counts: 2832 16003 6590 1136 1187
## probabilities: 0.102 0.577 0.237 0.041 0.043
## left son=302 (3747 obs) right son=303 (24001 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=743.1407, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=644.6047, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=571.9225, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=471.3980, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=458.7793, (0 missing)
##
## Node number 158: 361 observations
## predicted class=Employed expected loss=0.09418283 P(node) =0.0046795
## class counts: 1 327 0 5 28
## probabilities: 0.003 0.906 0.000 0.014 0.078
##
## Node number 159: 3877 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5870518 P(node) =0.05025601
## class counts: 635 1290 284 1601 67
## probabilities: 0.164 0.333 0.073 0.413 0.017
## left son=318 (242 obs) right son=319 (3635 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=152.6798, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=131.5716, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=108.9098, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=106.1004, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=102.9366, (0 missing)
##
## Node number 190: 61 observations
## predicted class=Employed expected loss=0.3606557 P(node) =0.0007907188
## class counts: 0 39 9 0 13
## probabilities: 0.000 0.639 0.148 0.000 0.213
##
## Node number 191: 4072 observations, complexity param=0.0009684776
## predicted class=Not.in.Labor.Force expected loss=0.08423379 P(node) =0.05278372
## class counts: 60 131 3729 2 150
## probabilities: 0.015 0.032 0.916 0.000 0.037
## left son=382 (42 obs) right son=383 (4030 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=58.11483, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=43.22076, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=39.39532, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=28.43424, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=24.96236, (0 missing)
##
## Node number 254: 222 observations
## predicted class=Employed expected loss=0.07207207 P(node) =0.002877698
## class counts: 0 206 1 7 8
## probabilities: 0.000 0.928 0.005 0.032 0.036
##
## Node number 255: 12572 observations, complexity param=0.006092037
## predicted class=Retired expected loss=0.1420617 P(node) =0.1629658
## class counts: 623 795 324 10786 44
## probabilities: 0.050 0.063 0.026 0.858 0.003
## left son=510 (215 obs) right son=511 (12357 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=317.9065, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=245.9525, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=212.9354, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=198.9882, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=198.2375, (0 missing)
##
## Node number 302: 3747 observations
## predicted class=Employed expected loss=0.09634374 P(node) =0.04857087
## class counts: 2 3386 43 3 313
## probabilities: 0.001 0.904 0.011 0.001 0.084
##
## Node number 303: 24001 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.4743136 P(node) =0.3111154
## class counts: 2830 12617 6547 1133 874
## probabilities: 0.118 0.526 0.273 0.047 0.036
## left son=606 (2696 obs) right son=607 (21305 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=849.2295, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=771.8555, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=616.0829, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=606.5147, (0 missing)
## Age < 26.5 to the right, improve=546.5604, (0 missing)
##
## Node number 318: 242 observations
## predicted class=Employed expected loss=0.04132231 P(node) =0.00313695
## class counts: 0 232 0 1 9
## probabilities: 0.000 0.959 0.000 0.004 0.037
##
## Node number 319: 3635 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5598349 P(node) =0.04711906
## class counts: 635 1058 284 1600 58
## probabilities: 0.175 0.291 0.078 0.440 0.016
## left son=638 (226 obs) right son=639 (3409 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=151.0187, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=125.3594, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=121.5617, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=117.7939, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=108.4572, (0 missing)
##
## Node number 382: 42 observations
## predicted class=Employed expected loss=0.1904762 P(node) =0.0005444293
## class counts: 0 34 2 0 6
## probabilities: 0.000 0.810 0.048 0.000 0.143
##
## Node number 383: 4030 observations, complexity param=0.0007497891
## predicted class=Not.in.Labor.Force expected loss=0.0751861 P(node) =0.05223929
## class counts: 60 97 3727 2 144
## probabilities: 0.015 0.024 0.925 0.000 0.036
## left son=766 (30 obs) right son=767 (4000 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=44.11017, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=40.32177, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=28.97328, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=25.41738, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=21.77548, (0 missing)
##
## Node number 510: 215 observations
## predicted class=Employed expected loss=0.06976744 P(node) =0.00278696
## class counts: 0 200 0 5 10
## probabilities: 0.000 0.930 0.000 0.023 0.047
##
## Node number 511: 12357 observations, complexity param=0.004717423
## predicted class=Retired expected loss=0.127539 P(node) =0.1601789
## class counts: 623 595 324 10781 34
## probabilities: 0.050 0.048 0.026 0.872 0.003
## left son=1022 (158 obs) right son=1023 (12199 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=254.39830, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=220.42630, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=206.10560, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=205.04780, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 80.04601, (0 missing)
##
## Node number 606: 2696 observations
## predicted class=Employed expected loss=0.03783383 P(node) =0.03494718
## class counts: 0 2594 7 1 94
## probabilities: 0.000 0.962 0.003 0.000 0.035
##
## Node number 607: 21305 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.5295471 P(node) =0.2761683
## class counts: 2830 10023 6540 1132 780
## probabilities: 0.133 0.470 0.307 0.053 0.037
## left son=1214 (2819 obs) right son=1215 (18486 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=1012.4120, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 789.2268, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 783.2167, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 680.7946, (0 missing)
## Age < 26.5 to the right, improve= 519.2679, (0 missing)
##
## Node number 638: 226 observations
## predicted class=Employed expected loss=0.0619469 P(node) =0.002929548
## class counts: 0 212 0 6 8
## probabilities: 0.000 0.938 0.000 0.027 0.035
##
## Node number 639: 3409 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5324142 P(node) =0.04418951
## class counts: 635 846 284 1594 50
## probabilities: 0.186 0.248 0.083 0.468 0.015
## left son=1278 (199 obs) right son=1279 (3210 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=143.50720, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=138.57380, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=134.13540, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=123.53620, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 76.56807, (0 missing)
##
## Node number 766: 30 observations
## predicted class=Employed expected loss=0.1666667 P(node) =0.0003888781
## class counts: 0 25 1 0 4
## probabilities: 0.000 0.833 0.033 0.000 0.133
##
## Node number 767: 4000 observations, complexity param=0.0006248243
## predicted class=Not.in.Labor.Force expected loss=0.0685 P(node) =0.05185041
## class counts: 60 72 3726 2 140
## probabilities: 0.015 0.018 0.931 0.001 0.035
## left son=1534 (36 obs) right son=1535 (3964 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=41.016160, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=29.377430, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=25.758690, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.067800, (0 missing)
## Age < 17.5 to the left, improve= 9.052118, (0 missing)
##
## Node number 1022: 158 observations
## predicted class=Employed expected loss=0.03164557 P(node) =0.002048091
## class counts: 0 153 0 2 3
## probabilities: 0.000 0.968 0.000 0.013 0.019
##
## Node number 1023: 12199 observations, complexity param=0.004092599
## predicted class=Retired expected loss=0.116403 P(node) =0.1581308
## class counts: 623 442 324 10779 31
## probabilities: 0.051 0.036 0.027 0.884 0.003
## left son=2046 (144 obs) right son=2047 (12055 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=226.33950, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=211.73010, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=210.42740, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 82.21373, (0 missing)
## Age < 69.5 to the left, improve= 54.47671, (0 missing)
##
## Node number 1214: 2819 observations
## predicted class=Employed expected loss=0.08300816 P(node) =0.03654158
## class counts: 2 2585 17 3 212
## probabilities: 0.001 0.917 0.006 0.001 0.075
##
## Node number 1215: 18486 observations, complexity param=0.009256772
## predicted class=Employed expected loss=0.5976415 P(node) =0.2396267
## class counts: 2828 7438 6523 1129 568
## probabilities: 0.153 0.402 0.353 0.061 0.031
## left son=2430 (2078 obs) right son=2431 (16408 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=1045.4260, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1044.0940, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 916.3636, (0 missing)
## Age < 26.5 to the right, improve= 504.6234, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 382.4998, (0 missing)
##
## Node number 1278: 199 observations
## predicted class=Employed expected loss=0.09547739 P(node) =0.002579558
## class counts: 1 180 0 3 15
## probabilities: 0.005 0.905 0.000 0.015 0.075
##
## Node number 1279: 3210 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.5043614 P(node) =0.04160996
## class counts: 634 666 284 1591 35
## probabilities: 0.198 0.207 0.088 0.496 0.011
## left son=2558 (182 obs) right son=2559 (3028 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=156.29700, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=151.07500, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=139.20960, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 86.10358, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 47.69273, (0 missing)
##
## Node number 1534: 36 observations
## predicted class=Employed expected loss=0.3333333 P(node) =0.0004666537
## class counts: 0 24 4 0 8
## probabilities: 0.000 0.667 0.111 0.000 0.222
##
## Node number 1535: 3964 observations, complexity param=0.0004998594
## predicted class=Not.in.Labor.Force expected loss=0.06104945 P(node) =0.05138376
## class counts: 60 48 3722 2 132
## probabilities: 0.015 0.012 0.939 0.001 0.033
## left son=3070 (17 obs) right son=3071 (3947 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=29.804300, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=26.116770, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.374470, (0 missing)
## Age < 17.5 to the left, improve= 6.419004, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.571355, (0 missing)
##
## Node number 2046: 144 observations
## predicted class=Employed expected loss=0.09027778 P(node) =0.001866615
## class counts: 0 131 1 0 12
## probabilities: 0.000 0.910 0.007 0.000 0.083
##
## Node number 2047: 12055 observations, complexity param=0.003842669
## predicted class=Retired expected loss=0.1058482 P(node) =0.1562642
## class counts: 623 311 323 10779 19
## probabilities: 0.052 0.026 0.027 0.894 0.002
## left son=4094 (139 obs) right son=4095 (11916 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=216.92580, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=215.39200, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 84.21834, (0 missing)
## Age < 69.5 to the left, improve= 42.58446, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 21.11926, (0 missing)
##
## Node number 2430: 2078 observations
## predicted class=Employed expected loss=0.05293551 P(node) =0.02693629
## class counts: 0 1968 6 2 102
## probabilities: 0.000 0.947 0.003 0.001 0.049
##
## Node number 2431: 16408 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.6028157 P(node) =0.2126904
## class counts: 2828 5470 6517 1127 466
## probabilities: 0.172 0.333 0.397 0.069 0.028
## left son=4862 (1984 obs) right son=4863 (14424 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1338.0430, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=1188.9860, (0 missing)
## Age < 41.5 to the right, improve= 523.3124, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 491.6569, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 358.8742, (0 missing)
##
## Node number 2558: 182 observations
## predicted class=Employed expected loss=0.07142857 P(node) =0.002359194
## class counts: 0 169 0 1 12
## probabilities: 0.000 0.929 0.000 0.005 0.066
##
## Node number 2559: 3028 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.4749009 P(node) =0.03925076
## class counts: 634 497 284 1590 23
## probabilities: 0.209 0.164 0.094 0.525 0.008
## left son=5118 (174 obs) right son=5119 (2854 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=170.19160, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=156.88020, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 96.83817, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 53.76955, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 30.07375, (0 missing)
##
## Node number 3070: 17 observations
## predicted class=Employed expected loss=0.05882353 P(node) =0.0002203642
## class counts: 0 16 0 0 1
## probabilities: 0.000 0.941 0.000 0.000 0.059
##
## Node number 3071: 3947 observations, complexity param=0.000437377
## predicted class=Not.in.Labor.Force expected loss=0.05700532 P(node) =0.05116339
## class counts: 60 32 3722 2 131
## probabilities: 0.015 0.008 0.943 0.001 0.033
## left son=6142 (14 obs) right son=6143 (3933 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=26.335330, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=22.561670, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.647505, (0 missing)
## Age < 17.5 to the left, improve= 4.420919, (0 missing)
##
## Node number 4094: 139 observations
## predicted class=Employed expected loss=0.08633094 P(node) =0.001801802
## class counts: 0 127 1 4 7
## probabilities: 0.000 0.914 0.007 0.029 0.050
##
## Node number 4095: 11916 observations, complexity param=0.003811428
## predicted class=Retired expected loss=0.09575361 P(node) =0.1544624
## class counts: 623 184 322 10775 12
## probabilities: 0.052 0.015 0.027 0.904 0.001
## left son=8190 (128 obs) right son=8191 (11788 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=220.28360, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 86.19226, (0 missing)
## Age < 69.5 to the left, improve= 32.25519, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 21.65488, (0 missing)
##
## Node number 4862: 1984 observations
## predicted class=Employed expected loss=0.03175403 P(node) =0.0257178
## class counts: 1 1921 5 3 54
## probabilities: 0.001 0.968 0.003 0.002 0.027
##
## Node number 4863: 14424 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.5485302 P(node) =0.1869726
## class counts: 2827 3549 6512 1124 412
## probabilities: 0.196 0.246 0.451 0.078 0.029
## left son=9726 (1952 obs) right son=9727 (12472 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=1586.3880, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 648.4297, (0 missing)
## Age < 41.5 to the right, improve= 548.0990, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 472.6345, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 283.5885, (0 missing)
##
## Node number 5118: 174 observations
## predicted class=Employed expected loss=0.04597701 P(node) =0.002255493
## class counts: 0 166 1 5 2
## probabilities: 0.000 0.954 0.006 0.029 0.011
##
## Node number 5119: 2854 observations, complexity param=0.003238672
## predicted class=Retired expected loss=0.4446391 P(node) =0.03699527
## class counts: 634 331 283 1585 21
## probabilities: 0.222 0.116 0.099 0.555 0.007
## left son=10238 (162 obs) right son=10239 (2692 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=177.24140, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=109.19970, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 60.72953, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 33.66763, (0 missing)
## Age < 61.5 to the left, improve= 24.28196, (0 missing)
##
## Node number 6142: 14 observations
## predicted class=Employed expected loss=0 P(node) =0.0001814764
## class counts: 0 14 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 6143: 3933 observations, complexity param=0.0003748946
## predicted class=Not.in.Labor.Force expected loss=0.05364861 P(node) =0.05098192
## class counts: 60 18 3722 2 131
## probabilities: 0.015 0.005 0.946 0.001 0.033
## left son=12286 (12 obs) right son=12287 (3921 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=22.722820, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 5.713301, (0 missing)
## Age < 17.5 to the left, improve= 3.677531, (0 missing)
##
## Node number 8190: 128 observations
## predicted class=Employed expected loss=0.046875 P(node) =0.001659213
## class counts: 0 122 0 0 6
## probabilities: 0.000 0.953 0.000 0.000 0.047
##
## Node number 8191: 11788 observations, complexity param=0.001499578
## predicted class=Retired expected loss=0.08593485 P(node) =0.1528032
## class counts: 623 62 322 10775 6
## probabilities: 0.053 0.005 0.027 0.914 0.001
## left son=16382 (54 obs) right son=16383 (11734 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=88.14437, (0 missing)
## Age < 66.5 to the left, improve=24.62345, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=22.18490, (0 missing)
##
## Node number 9726: 1952 observations
## predicted class=Employed expected loss=0.07633197 P(node) =0.025303
## class counts: 2 1803 13 0 134
## probabilities: 0.001 0.924 0.007 0.000 0.069
##
## Node number 9727: 12472 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.4789128 P(node) =0.1616696
## class counts: 2825 1746 6499 1124 278
## probabilities: 0.227 0.140 0.521 0.090 0.022
## left son=19454 (864 obs) right son=19455 (11608 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=875.54450, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=636.69090, (0 missing)
## Age < 44.5 to the right, improve=628.14260, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=376.68880, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 25.43837, (0 missing)
##
## Node number 10238: 162 observations
## predicted class=Employed expected loss=0.0617284 P(node) =0.002099942
## class counts: 0 152 0 2 8
## probabilities: 0.000 0.938 0.000 0.012 0.049
##
## Node number 10239: 2692 observations, complexity param=0.002936674
## predicted class=Retired expected loss=0.4119614 P(node) =0.03489533
## class counts: 634 179 283 1583 13
## probabilities: 0.236 0.066 0.105 0.588 0.005
## left son=20478 (101 obs) right son=20479 (2591 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=122.93260, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 68.49246, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 37.63462, (0 missing)
## Age < 61.5 to the left, improve= 23.23422, (0 missing)
##
## Node number 12286: 12 observations
## predicted class=Employed expected loss=0 P(node) =0.0001555512
## class counts: 0 12 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 12287: 3921 observations, complexity param=6.248243e-05
## predicted class=Not.in.Labor.Force expected loss=0.05075236 P(node) =0.05082637
## class counts: 60 6 3722 2 131
## probabilities: 0.015 0.002 0.949 0.001 0.033
## left son=24574 (8 obs) right son=24575 (3913 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=5.770378, (0 missing)
## Age < 17.5 to the left, improve=3.474435, (0 missing)
##
## Node number 16382: 54 observations
## predicted class=Employed expected loss=0.09259259 P(node) =0.0006999806
## class counts: 0 49 0 1 4
## probabilities: 0.000 0.907 0.000 0.019 0.074
##
## Node number 16383: 11734 observations, complexity param=0.0001718267
## predicted class=Retired expected loss=0.08181353 P(node) =0.1521032
## class counts: 623 13 322 10774 2
## probabilities: 0.053 0.001 0.027 0.918 0.000
## left son=32766 (1428 obs) right son=32767 (10306 obs)
## Primary splits:
## Age < 66.5 to the left, improve=22.59645, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=22.40638, (0 missing)
##
## Node number 19454: 864 observations
## predicted class=Employed expected loss=0.07407407 P(node) =0.01119969
## class counts: 0 800 6 1 57
## probabilities: 0.000 0.926 0.007 0.001 0.066
##
## Node number 19455: 11608 observations, complexity param=0.009256772
## predicted class=Not.in.Labor.Force expected loss=0.4406444 P(node) =0.1504699
## class counts: 2825 946 6493 1123 221
## probabilities: 0.243 0.081 0.559 0.097 0.019
## left son=38910 (642 obs) right son=38911 (10966 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=738.10700, (0 missing)
## Age < 44.5 to the right, improve=681.98090, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=433.82320, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 26.34573, (0 missing)
##
## Node number 20478: 101 observations
## predicted class=Employed expected loss=0.04950495 P(node) =0.001309223
## class counts: 0 96 0 2 3
## probabilities: 0.000 0.950 0.000 0.020 0.030
##
## Node number 20479: 2591 observations, complexity param=0.001687026
## predicted class=Retired expected loss=0.3898109 P(node) =0.0335861
## class counts: 634 83 283 1581 10
## probabilities: 0.245 0.032 0.109 0.610 0.004
## left son=40958 (60 obs) right son=40959 (2531 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=74.12104, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=40.50551, (0 missing)
## Age < 61.5 to the left, improve=22.94488, (0 missing)
##
## Node number 24574: 8 observations
## predicted class=Employed expected loss=0.375 P(node) =0.0001037008
## class counts: 0 5 3 0 0
## probabilities: 0.000 0.625 0.375 0.000 0.000
##
## Node number 24575: 3913 observations
## predicted class=Not.in.Labor.Force expected loss=0.04957833 P(node) =0.05072267
## class counts: 60 1 3719 2 131
## probabilities: 0.015 0.000 0.950 0.001 0.033
##
## Node number 32766: 1428 observations
## predicted class=Retired expected loss=0.17507 P(node) =0.0185106
## class counts: 172 2 76 1178 0
## probabilities: 0.120 0.001 0.053 0.825 0.000
##
## Node number 32767: 10306 observations, complexity param=0.0001718267
## predicted class=Retired expected loss=0.06889191 P(node) =0.1335926
## class counts: 451 11 246 9596 2
## probabilities: 0.044 0.001 0.024 0.931 0.000
## left son=65534 (12 obs) right son=65535 (10294 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=20.600160, (0 missing)
## Age < 70.5 to the left, improve= 4.406275, (0 missing)
##
## Node number 38910: 642 observations
## predicted class=Employed expected loss=0.07165109 P(node) =0.008321991
## class counts: 2 596 7 1 36
## probabilities: 0.003 0.928 0.011 0.002 0.056
##
## Node number 38911: 10966 observations, complexity param=0.007591615
## predicted class=Not.in.Labor.Force expected loss=0.4085355 P(node) =0.1421479
## class counts: 2823 350 6486 1122 185
## probabilities: 0.257 0.032 0.591 0.102 0.017
## left son=77822 (4676 obs) right son=77823 (6290 obs)
## Primary splits:
## Age < 43.5 to the right, improve=710.94140, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=485.31870, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 27.25333, (0 missing)
##
## Node number 40958: 60 observations
## predicted class=Employed expected loss=0.1 P(node) =0.0007777562
## class counts: 0 54 0 0 6
## probabilities: 0.000 0.900 0.000 0.000 0.100
##
## Node number 40959: 2531 observations, complexity param=0.0009059952
## predicted class=Retired expected loss=0.3753457 P(node) =0.03280835
## class counts: 634 29 283 1581 4
## probabilities: 0.250 0.011 0.112 0.625 0.002
## left son=81918 (29 obs) right son=81919 (2502 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=42.32207, (0 missing)
## Age < 61.5 to the left, improve=22.59598, (0 missing)
##
## Node number 65534: 12 observations
## predicted class=Employed expected loss=0.08333333 P(node) =0.0001555512
## class counts: 0 11 0 0 1
## probabilities: 0.000 0.917 0.000 0.000 0.083
##
## Node number 65535: 10294 observations
## predicted class=Retired expected loss=0.06780649 P(node) =0.133437
## class counts: 451 0 246 9596 1
## probabilities: 0.044 0.000 0.024 0.932 0.000
##
## Node number 77822: 4676 observations, complexity param=0.004592458
## predicted class=Disabled expected loss=0.5902481 P(node) =0.06061313
## class counts: 1916 147 1673 927 13
## probabilities: 0.410 0.031 0.358 0.198 0.003
## left son=155644 (152 obs) right son=155645 (4524 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=190.3492, (0 missing)
## Age < 53.5 to the left, improve= 92.3401, (0 missing)
##
## Node number 77823: 6290 observations, complexity param=0.006310725
## predicted class=Not.in.Labor.Force expected loss=0.2348172 P(node) =0.08153477
## class counts: 907 203 4813 195 172
## probabilities: 0.144 0.032 0.765 0.031 0.027
## left son=155646 (211 obs) right son=155647 (6079 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=319.75960, (0 missing)
## Age < 26.5 to the right, improve= 58.87956, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 29.62228, (0 missing)
##
## Node number 81918: 29 observations
## predicted class=Employed expected loss=0 P(node) =0.0003759155
## class counts: 0 29 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 81919: 2502 observations
## predicted class=Retired expected loss=0.3681055 P(node) =0.03243243
## class counts: 634 0 283 1581 4
## probabilities: 0.253 0.000 0.113 0.632 0.002
##
## Node number 155644: 152 observations
## predicted class=Employed expected loss=0.03289474 P(node) =0.001970316
## class counts: 0 147 0 0 5
## probabilities: 0.000 0.967 0.000 0.000 0.033
##
## Node number 155645: 4524 observations, complexity param=0.003249086
## predicted class=Disabled expected loss=0.576481 P(node) =0.05864282
## class counts: 1916 0 1673 927 8
## probabilities: 0.424 0.000 0.370 0.205 0.002
## left son=311290 (2474 obs) right son=311291 (2050 obs)
## Primary splits:
## Age < 53.5 to the left, improve=96.05737, (0 missing)
##
## Node number 155646: 211 observations
## predicted class=Employed expected loss=0.03791469 P(node) =0.002735109
## class counts: 0 203 1 0 7
## probabilities: 0.000 0.962 0.005 0.000 0.033
##
## Node number 155647: 6079 observations, complexity param=0.0002030679
## predicted class=Not.in.Labor.Force expected loss=0.2084224 P(node) =0.07879966
## class counts: 907 0 4812 195 165
## probabilities: 0.149 0.000 0.792 0.032 0.027
## left son=311294 (3558 obs) right son=311295 (2521 obs)
## Primary splits:
## Age < 26.5 to the right, improve=51.65542, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve=30.42549, (0 missing)
##
## Node number 311290: 2474 observations, complexity param=0.002530538
## predicted class=Not.in.Labor.Force expected loss=0.5359741 P(node) =0.03206948
## class counts: 1044 0 1148 276 6
## probabilities: 0.422 0.000 0.464 0.112 0.002
## left son=622580 (1136 obs) right son=622581 (1338 obs)
## Primary splits:
## Age < 49.5 to the right, improve=20.27439, (0 missing)
##
## Node number 311291: 2050 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.5746341 P(node) =0.02657334
## class counts: 872 0 525 651 2
## probabilities: 0.425 0.000 0.256 0.318 0.001
## left son=622582 (987 obs) right son=622583 (1063 obs)
## Primary splits:
## Age < 56.5 to the left, improve=11.09287, (0 missing)
##
## Node number 311294: 3558 observations, complexity param=0.0002030679
## predicted class=Not.in.Labor.Force expected loss=0.2613828 P(node) =0.04612094
## class counts: 724 0 2628 153 53
## probabilities: 0.203 0.000 0.739 0.043 0.015
## left son=622588 (3545 obs) right son=622589 (13 obs)
## Primary splits:
## Industry.my.fctrArmed forces < 0.5 to the left, improve=20.34446, (0 missing)
## Age < 37.5 to the right, improve=13.39218, (0 missing)
##
## Node number 311295: 2521 observations
## predicted class=Not.in.Labor.Force expected loss=0.1336771 P(node) =0.03267872
## class counts: 183 0 2184 42 112
## probabilities: 0.073 0.000 0.866 0.017 0.044
##
## Node number 622580: 1136 observations
## predicted class=Disabled expected loss=0.5440141 P(node) =0.01472552
## class counts: 518 0 437 180 1
## probabilities: 0.456 0.000 0.385 0.158 0.001
##
## Node number 622581: 1338 observations
## predicted class=Not.in.Labor.Force expected loss=0.4686099 P(node) =0.01734396
## class counts: 526 0 711 96 5
## probabilities: 0.393 0.000 0.531 0.072 0.004
##
## Node number 622582: 987 observations
## predicted class=Disabled expected loss=0.5542047 P(node) =0.01279409
## class counts: 440 0 293 253 1
## probabilities: 0.446 0.000 0.297 0.256 0.001
##
## Node number 622583: 1063 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.593603 P(node) =0.01377925
## class counts: 432 0 232 398 1
## probabilities: 0.406 0.000 0.218 0.374 0.001
## left son=1245166 (314 obs) right son=1245167 (749 obs)
## Primary splits:
## Age < 57.5 to the left, improve=1.268934, (0 missing)
##
## Node number 622588: 3545 observations
## predicted class=Not.in.Labor.Force expected loss=0.2586742 P(node) =0.04595243
## class counts: 724 0 2628 153 40
## probabilities: 0.204 0.000 0.741 0.043 0.011
##
## Node number 622589: 13 observations
## predicted class=Unemployed expected loss=0 P(node) =0.0001685138
## class counts: 0 0 0 0 13
## probabilities: 0.000 0.000 0.000 0.000 1.000
##
## Node number 1245166: 314 observations
## predicted class=Disabled expected loss=0.5732484 P(node) =0.004070257
## class counts: 134 0 76 104 0
## probabilities: 0.427 0.000 0.242 0.331 0.000
##
## Node number 1245167: 749 observations, complexity param=3.124121e-05
## predicted class=Disabled expected loss=0.6021362 P(node) =0.00970899
## class counts: 298 0 156 294 1
## probabilities: 0.398 0.000 0.208 0.393 0.001
## left son=2490334 (385 obs) right son=2490335 (364 obs)
## Primary splits:
## Age < 58.5 to the left, improve=0.1735246, (0 missing)
##
## Node number 2490334: 385 observations
## predicted class=Disabled expected loss=0.6 P(node) =0.004990602
## class counts: 154 0 84 147 0
## probabilities: 0.400 0.000 0.218 0.382 0.000
##
## Node number 2490335: 364 observations
## predicted class=Retired expected loss=0.5961538 P(node) =0.004718387
## class counts: 144 0 72 147 1
## probabilities: 0.396 0.000 0.198 0.404 0.003
##
## n= 77145
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 77145 32009 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 64.5 62433 20075 Employed (0.057 0.68 0.17 0.044 0.047)
## 4) Age>=18.5 57429 15973 Employed (0.061 0.72 0.12 0.048 0.047)
## 8) Industry.my.fctrEducational and health services>=0.5 10137 508 Employed (0.00089 0.95 0.005 0.0013 0.043) *
## 9) Industry.my.fctrEducational and health services< 0.5 47292 15465 Employed (0.074 0.67 0.15 0.058 0.048)
## 18) Age< 59.5 42249 12793 Employed (0.067 0.7 0.16 0.027 0.05)
## 36) Industry.my.fctrTrade>=0.5 5469 432 Employed (0.0013 0.92 0.0073 0.00018 0.07) *
## 37) Industry.my.fctrTrade< 0.5 36780 12361 Employed (0.077 0.66 0.18 0.031 0.048)
## 74) Industry.my.fctrProfessional and business services>=0.5 4668 337 Employed (0.00064 0.93 0.0066 0 0.065) *
## 75) Industry.my.fctrProfessional and business services< 0.5 32112 12024 Employed (0.088 0.63 0.21 0.035 0.045)
## 150) Industry.my.fctrManufacturing>=0.5 4364 279 Employed (0.00046 0.94 0.0034 0.00069 0.059) *
## 151) Industry.my.fctrManufacturing< 0.5 27748 11745 Employed (0.1 0.58 0.24 0.041 0.043)
## 302) Industry.my.fctrLeisure and hospitality>=0.5 3747 361 Employed (0.00053 0.9 0.011 0.0008 0.084) *
## 303) Industry.my.fctrLeisure and hospitality< 0.5 24001 11384 Employed (0.12 0.53 0.27 0.047 0.036)
## 606) Industry.my.fctrFinancial>=0.5 2696 102 Employed (0 0.96 0.0026 0.00037 0.035) *
## 607) Industry.my.fctrFinancial< 0.5 21305 11282 Employed (0.13 0.47 0.31 0.053 0.037)
## 1214) Industry.my.fctrConstruction>=0.5 2819 234 Employed (0.00071 0.92 0.006 0.0011 0.075) *
## 1215) Industry.my.fctrConstruction< 0.5 18486 11048 Employed (0.15 0.4 0.35 0.061 0.031)
## 2430) Industry.my.fctrTransportation and utilities>=0.5 2078 110 Employed (0 0.95 0.0029 0.00096 0.049) *
## 2431) Industry.my.fctrTransportation and utilities< 0.5 16408 9891 Not.in.Labor.Force (0.17 0.33 0.4 0.069 0.028)
## 4862) Industry.my.fctrPublic administration>=0.5 1984 63 Employed (0.0005 0.97 0.0025 0.0015 0.027) *
## 4863) Industry.my.fctrPublic administration< 0.5 14424 7912 Not.in.Labor.Force (0.2 0.25 0.45 0.078 0.029)
## 9726) Industry.my.fctrOther services>=0.5 1952 149 Employed (0.001 0.92 0.0067 0 0.069) *
## 9727) Industry.my.fctrOther services< 0.5 12472 5973 Not.in.Labor.Force (0.23 0.14 0.52 0.09 0.022)
## 19454) Industry.my.fctrInformation>=0.5 864 64 Employed (0 0.93 0.0069 0.0012 0.066) *
## 19455) Industry.my.fctrInformation< 0.5 11608 5115 Not.in.Labor.Force (0.24 0.081 0.56 0.097 0.019)
## 38910) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 642 46 Employed (0.0031 0.93 0.011 0.0016 0.056) *
## 38911) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 10966 4480 Not.in.Labor.Force (0.26 0.032 0.59 0.1 0.017)
## 77822) Age>=43.5 4676 2760 Disabled (0.41 0.031 0.36 0.2 0.0028)
## 155644) Industry.my.fctrMining>=0.5 152 5 Employed (0 0.97 0 0 0.033) *
## 155645) Industry.my.fctrMining< 0.5 4524 2608 Disabled (0.42 0 0.37 0.2 0.0018)
## 311290) Age< 53.5 2474 1326 Not.in.Labor.Force (0.42 0 0.46 0.11 0.0024)
## 622580) Age>=49.5 1136 618 Disabled (0.46 0 0.38 0.16 0.00088) *
## 622581) Age< 49.5 1338 627 Not.in.Labor.Force (0.39 0 0.53 0.072 0.0037) *
## 311291) Age>=53.5 2050 1178 Disabled (0.43 0 0.26 0.32 0.00098)
## 622582) Age< 56.5 987 547 Disabled (0.45 0 0.3 0.26 0.001) *
## 622583) Age>=56.5 1063 631 Disabled (0.41 0 0.22 0.37 0.00094)
## 1245166) Age< 57.5 314 180 Disabled (0.43 0 0.24 0.33 0) *
## 1245167) Age>=57.5 749 451 Disabled (0.4 0 0.21 0.39 0.0013)
## 2490334) Age< 58.5 385 231 Disabled (0.4 0 0.22 0.38 0) *
## 2490335) Age>=58.5 364 217 Retired (0.4 0 0.2 0.4 0.0027) *
## 77823) Age< 43.5 6290 1477 Not.in.Labor.Force (0.14 0.032 0.77 0.031 0.027)
## 155646) Industry.my.fctrMining>=0.5 211 8 Employed (0 0.96 0.0047 0 0.033) *
## 155647) Industry.my.fctrMining< 0.5 6079 1267 Not.in.Labor.Force (0.15 0 0.79 0.032 0.027)
## 311294) Age>=26.5 3558 930 Not.in.Labor.Force (0.2 0 0.74 0.043 0.015)
## 622588) Industry.my.fctrArmed forces< 0.5 3545 917 Not.in.Labor.Force (0.2 0 0.74 0.043 0.011) *
## 622589) Industry.my.fctrArmed forces>=0.5 13 0 Unemployed (0 0 0 0 1) *
## 311295) Age< 26.5 2521 337 Not.in.Labor.Force (0.073 0 0.87 0.017 0.044) *
## 19) Age>=59.5 5043 2672 Employed (0.13 0.47 0.057 0.32 0.027)
## 38) Industry.my.fctrTrade>=0.5 451 25 Employed (0.0022 0.94 0.0022 0.0067 0.044) *
## 39) Industry.my.fctrTrade< 0.5 4592 2647 Employed (0.14 0.42 0.062 0.35 0.026)
## 78) Industry.my.fctrProfessional and business services>=0.5 354 26 Employed (0 0.93 0 0.0085 0.065) *
## 79) Industry.my.fctrProfessional and business services< 0.5 4238 2621 Employed (0.15 0.38 0.067 0.38 0.022)
## 158) Industry.my.fctrManufacturing>=0.5 361 34 Employed (0.0028 0.91 0 0.014 0.078) *
## 159) Industry.my.fctrManufacturing< 0.5 3877 2276 Retired (0.16 0.33 0.073 0.41 0.017)
## 318) Industry.my.fctrFinancial>=0.5 242 10 Employed (0 0.96 0 0.0041 0.037) *
## 319) Industry.my.fctrFinancial< 0.5 3635 2035 Retired (0.17 0.29 0.078 0.44 0.016)
## 638) Industry.my.fctrConstruction>=0.5 226 14 Employed (0 0.94 0 0.027 0.035) *
## 639) Industry.my.fctrConstruction< 0.5 3409 1815 Retired (0.19 0.25 0.083 0.47 0.015)
## 1278) Industry.my.fctrLeisure and hospitality>=0.5 199 19 Employed (0.005 0.9 0 0.015 0.075) *
## 1279) Industry.my.fctrLeisure and hospitality< 0.5 3210 1619 Retired (0.2 0.21 0.088 0.5 0.011)
## 2558) Industry.my.fctrTransportation and utilities>=0.5 182 13 Employed (0 0.93 0 0.0055 0.066) *
## 2559) Industry.my.fctrTransportation and utilities< 0.5 3028 1438 Retired (0.21 0.16 0.094 0.53 0.0076)
## 5118) Industry.my.fctrPublic administration>=0.5 174 8 Employed (0 0.95 0.0057 0.029 0.011) *
## 5119) Industry.my.fctrPublic administration< 0.5 2854 1269 Retired (0.22 0.12 0.099 0.56 0.0074)
## 10238) Industry.my.fctrOther services>=0.5 162 10 Employed (0 0.94 0 0.012 0.049) *
## 10239) Industry.my.fctrOther services< 0.5 2692 1109 Retired (0.24 0.066 0.11 0.59 0.0048)
## 20478) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 101 5 Employed (0 0.95 0 0.02 0.03) *
## 20479) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 2591 1010 Retired (0.24 0.032 0.11 0.61 0.0039)
## 40958) Industry.my.fctrInformation>=0.5 60 6 Employed (0 0.9 0 0 0.1) *
## 40959) Industry.my.fctrInformation< 0.5 2531 950 Retired (0.25 0.011 0.11 0.62 0.0016)
## 81918) Industry.my.fctrMining>=0.5 29 0 Employed (0 1 0 0 0) *
## 81919) Industry.my.fctrMining< 0.5 2502 921 Retired (0.25 0 0.11 0.63 0.0016) *
## 5) Age< 18.5 5004 1197 Not.in.Labor.Force (0.012 0.18 0.76 0.0004 0.047)
## 10) Industry.my.fctrLeisure and hospitality>=0.5 487 71 Employed (0 0.85 0.072 0 0.074) *
## 11) Industry.my.fctrLeisure and hospitality< 0.5 4517 745 Not.in.Labor.Force (0.013 0.11 0.84 0.00044 0.044)
## 22) Industry.my.fctrTrade>=0.5 232 31 Employed (0 0.87 0.034 0 0.099) *
## 23) Industry.my.fctrTrade< 0.5 4285 521 Not.in.Labor.Force (0.014 0.067 0.88 0.00047 0.041)
## 46) Industry.my.fctrEducational and health services>=0.5 97 24 Employed (0 0.75 0.18 0 0.072) *
## 47) Industry.my.fctrEducational and health services< 0.5 4188 441 Not.in.Labor.Force (0.014 0.051 0.89 0.00048 0.04)
## 94) Industry.my.fctrOther services>=0.5 55 13 Employed (0 0.76 0.16 0 0.073) *
## 95) Industry.my.fctrOther services< 0.5 4133 395 Not.in.Labor.Force (0.015 0.041 0.9 0.00048 0.039)
## 190) Industry.my.fctrProfessional and business services>=0.5 61 22 Employed (0 0.64 0.15 0 0.21) *
## 191) Industry.my.fctrProfessional and business services< 0.5 4072 343 Not.in.Labor.Force (0.015 0.032 0.92 0.00049 0.037)
## 382) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 42 8 Employed (0 0.81 0.048 0 0.14) *
## 383) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 4030 303 Not.in.Labor.Force (0.015 0.024 0.92 0.0005 0.036)
## 766) Industry.my.fctrConstruction>=0.5 30 5 Employed (0 0.83 0.033 0 0.13) *
## 767) Industry.my.fctrConstruction< 0.5 4000 274 Not.in.Labor.Force (0.015 0.018 0.93 0.0005 0.035)
## 1534) Industry.my.fctrManufacturing>=0.5 36 12 Employed (0 0.67 0.11 0 0.22) *
## 1535) Industry.my.fctrManufacturing< 0.5 3964 242 Not.in.Labor.Force (0.015 0.012 0.94 0.0005 0.033)
## 3070) Industry.my.fctrTransportation and utilities>=0.5 17 1 Employed (0 0.94 0 0 0.059) *
## 3071) Industry.my.fctrTransportation and utilities< 0.5 3947 225 Not.in.Labor.Force (0.015 0.0081 0.94 0.00051 0.033)
## 6142) Industry.my.fctrFinancial>=0.5 14 0 Employed (0 1 0 0 0) *
## 6143) Industry.my.fctrFinancial< 0.5 3933 211 Not.in.Labor.Force (0.015 0.0046 0.95 0.00051 0.033)
## 12286) Industry.my.fctrInformation>=0.5 12 0 Employed (0 1 0 0 0) *
## 12287) Industry.my.fctrInformation< 0.5 3921 199 Not.in.Labor.Force (0.015 0.0015 0.95 0.00051 0.033)
## 24574) Industry.my.fctrPublic administration>=0.5 8 3 Employed (0 0.63 0.37 0 0) *
## 24575) Industry.my.fctrPublic administration< 0.5 3913 194 Not.in.Labor.Force (0.015 0.00026 0.95 0.00051 0.033) *
## 3) Age>=64.5 14712 3866 Retired (0.043 0.19 0.022 0.74 0.0091)
## 6) Industry.my.fctrEducational and health services>=0.5 696 54 Employed (0 0.92 0 0.034 0.043) *
## 7) Industry.my.fctrEducational and health services< 0.5 14016 3194 Retired (0.045 0.15 0.023 0.77 0.0074)
## 14) Industry.my.fctrTrade>=0.5 402 27 Employed (0.005 0.93 0 0.03 0.032) *
## 15) Industry.my.fctrTrade< 0.5 13614 2804 Retired (0.046 0.13 0.024 0.79 0.0067)
## 30) Industry.my.fctrProfessional and business services>=0.5 384 33 Employed (0.0026 0.91 0.0052 0.026 0.052) *
## 31) Industry.my.fctrProfessional and business services< 0.5 13230 2430 Retired (0.047 0.11 0.025 0.82 0.0054)
## 62) Industry.my.fctrFinancial>=0.5 219 12 Employed (0 0.95 0 0.023 0.032) *
## 63) Industry.my.fctrFinancial< 0.5 13011 2216 Retired (0.048 0.092 0.025 0.83 0.0049)
## 126) Industry.my.fctrManufacturing>=0.5 217 15 Employed (0 0.93 0.0046 0.0092 0.055) *
## 127) Industry.my.fctrManufacturing< 0.5 12794 2001 Retired (0.049 0.078 0.025 0.84 0.0041)
## 254) Industry.my.fctrOther services>=0.5 222 16 Employed (0 0.93 0.0045 0.032 0.036) *
## 255) Industry.my.fctrOther services< 0.5 12572 1786 Retired (0.05 0.063 0.026 0.86 0.0035)
## 510) Industry.my.fctrLeisure and hospitality>=0.5 215 15 Employed (0 0.93 0 0.023 0.047) *
## 511) Industry.my.fctrLeisure and hospitality< 0.5 12357 1576 Retired (0.05 0.048 0.026 0.87 0.0028)
## 1022) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 158 5 Employed (0 0.97 0 0.013 0.019) *
## 1023) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 12199 1420 Retired (0.051 0.036 0.027 0.88 0.0025)
## 2046) Industry.my.fctrConstruction>=0.5 144 13 Employed (0 0.91 0.0069 0 0.083) *
## 2047) Industry.my.fctrConstruction< 0.5 12055 1276 Retired (0.052 0.026 0.027 0.89 0.0016)
## 4094) Industry.my.fctrTransportation and utilities>=0.5 139 12 Employed (0 0.91 0.0072 0.029 0.05) *
## 4095) Industry.my.fctrTransportation and utilities< 0.5 11916 1141 Retired (0.052 0.015 0.027 0.9 0.001)
## 8190) Industry.my.fctrPublic administration>=0.5 128 6 Employed (0 0.95 0 0 0.047) *
## 8191) Industry.my.fctrPublic administration< 0.5 11788 1013 Retired (0.053 0.0053 0.027 0.91 0.00051)
## 16382) Industry.my.fctrInformation>=0.5 54 5 Employed (0 0.91 0 0.019 0.074) *
## 16383) Industry.my.fctrInformation< 0.5 11734 960 Retired (0.053 0.0011 0.027 0.92 0.00017)
## 32766) Age< 66.5 1428 250 Retired (0.12 0.0014 0.053 0.82 0) *
## 32767) Age>=66.5 10306 710 Retired (0.044 0.0011 0.024 0.93 0.00019)
## 65534) Industry.my.fctrMining>=0.5 12 1 Employed (0 0.92 0 0 0.083) *
## 65535) Industry.my.fctrMining< 0.5 10294 698 Retired (0.044 0 0.024 0.93 9.7e-05) *## [1] TRUE# From here to save(), this should all be in one function
# these are executed in the same seq twice more:
# fit.data.training & predict.data.new chunks
glb_get_predictions <- function(df, mdl_id, rsp_var_out, prob_threshold_def=NULL) {
mdl <- glb_models_lst[[mdl_id]]
rsp_var_out <- paste0(rsp_var_out, mdl_id)
if (glb_is_regression) {
df[, rsp_var_out] <- predict(mdl, newdata=df, type="raw")
print(myplot_scatter(df, glb_rsp_var, rsp_var_out, smooth=TRUE))
df[, paste0(rsp_var_out, ".err")] <-
abs(df[, rsp_var_out] - df[, glb_rsp_var])
print(head(orderBy(reformulate(c("-", paste0(rsp_var_out, ".err"))),
df)))
}
if (glb_is_classification && glb_is_binomial) {
prob_threshold <- glb_models_df[glb_models_df$model_id == mdl_id,
"opt.prob.threshold.OOB"]
if (is.null(prob_threshold) || is.na(prob_threshold)) {
warning("Using default probability threshold: ", prob_threshold_def)
if (is.null(prob_threshold <- prob_threshold_def))
stop("Default probability threshold is NULL")
}
df[, paste0(rsp_var_out, ".prob")] <-
predict(mdl, newdata=df, type="prob")[, 2]
df[, rsp_var_out] <-
factor(levels(df[, glb_rsp_var])[
(df[, paste0(rsp_var_out, ".prob")] >=
prob_threshold) * 1 + 1], levels(df[, glb_rsp_var]))
# prediction stats already reported by myfit_mdl ???
}
if (glb_is_classification && !glb_is_binomial) {
df[, rsp_var_out] <- predict(mdl, newdata=df, type="raw")
df[, paste0(rsp_var_out, ".prob")] <-
predict(mdl, newdata=df, type="prob")
}
return(df)
}
glb_OOBobs_df <- glb_get_predictions(df=glb_OOBobs_df, mdl_id=glb_sel_mdl_id,
rsp_var_out=glb_rsp_var_out)## Warning in `[<-.data.frame`(`*tmp*`, , paste0(rsp_var_out, ".prob"), value
## = structure(list(: provided 5 variables to replace 1 variablespredct_accurate_var_name <- paste0(glb_rsp_var_out, glb_sel_mdl_id, ".accurate")
glb_OOBobs_df[, predct_accurate_var_name] <-
(glb_OOBobs_df[, glb_rsp_var] ==
glb_OOBobs_df[, paste0(glb_rsp_var_out, glb_sel_mdl_id)])
#stop(here"); sav_models_lst <- glb_models_lst; sav_models_df <- glb_models_df
glb_featsimp_df <-
myget_feats_importance(mdl=glb_sel_mdl, featsimp_df=NULL)
glb_featsimp_df[, paste0(glb_sel_mdl_id, ".importance")] <- glb_featsimp_df$importance
print(glb_featsimp_df)## importance
## Age 100.000000
## Industry.my.fctrOther services 33.090343
## Industry.my.fctrPublic administration 29.566192
## Industry.my.fctrManufacturing 28.370178
## Industry.my.fctrEducational and health services 27.058509
## Industry.my.fctrProfessional and business services 25.591436
## Industry.my.fctrTrade 24.329022
## Industry.my.fctrTransportation and utilities 23.012814
## Industry.my.fctrConstruction 22.685481
## Industry.my.fctrLeisure and hospitality 22.062655
## Industry.my.fctrFinancial 21.645346
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 20.237759
## Industry.my.fctrInformation 15.923595
## Industry.my.fctrMining 11.722476
## Industry.my.fctrArmed forces 0.784697
## `Industry.my.fctrLeisure and hospitality` 0.000000
## `Industry.my.fctrEducational and health services` 0.000000
## `Industry.my.fctrOther services` 0.000000
## `Industry.my.fctrTransportation and utilities` 0.000000
## `Industry.my.fctrProfessional and business services` 0.000000
## `Industry.my.fctrPublic administration` 0.000000
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.000000
## `Industry.my.fctrArmed forces` 0.000000
## Max.cor.Y.cv.0.cp.0.rpart.importance
## Age 100.000000
## Industry.my.fctrOther services 33.090343
## Industry.my.fctrPublic administration 29.566192
## Industry.my.fctrManufacturing 28.370178
## Industry.my.fctrEducational and health services 27.058509
## Industry.my.fctrProfessional and business services 25.591436
## Industry.my.fctrTrade 24.329022
## Industry.my.fctrTransportation and utilities 23.012814
## Industry.my.fctrConstruction 22.685481
## Industry.my.fctrLeisure and hospitality 22.062655
## Industry.my.fctrFinancial 21.645346
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 20.237759
## Industry.my.fctrInformation 15.923595
## Industry.my.fctrMining 11.722476
## Industry.my.fctrArmed forces 0.784697
## `Industry.my.fctrLeisure and hospitality` 0.000000
## `Industry.my.fctrEducational and health services` 0.000000
## `Industry.my.fctrOther services` 0.000000
## `Industry.my.fctrTransportation and utilities` 0.000000
## `Industry.my.fctrProfessional and business services` 0.000000
## `Industry.my.fctrPublic administration` 0.000000
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.000000
## `Industry.my.fctrArmed forces` 0.000000# Used again in fit.data.training & predict.data.new chunks
glb_analytics_diag_plots <- function(obs_df, mdl_id, prob_threshold=NULL) {
featsimp_df <- glb_featsimp_df
featsimp_df$feat <- gsub("`(.*?)`", "\\1", row.names(featsimp_df))
featsimp_df$feat.interact <- gsub("(.*?):(.*)", "\\2", featsimp_df$feat)
featsimp_df$feat <- gsub("(.*?):(.*)", "\\1", featsimp_df$feat)
featsimp_df$feat.interact <- ifelse(featsimp_df$feat.interact == featsimp_df$feat,
NA, featsimp_df$feat.interact)
featsimp_df$feat <- gsub("(.*?)\\.fctr(.*)", "\\1\\.fctr", featsimp_df$feat)
featsimp_df$feat.interact <- gsub("(.*?)\\.fctr(.*)", "\\1\\.fctr", featsimp_df$feat.interact)
featsimp_df <- orderBy(~ -importance.max, summaryBy(importance ~ feat + feat.interact,
data=featsimp_df, FUN=max))
#rex_str=":(.*)"; txt_vctr=tail(featsimp_df$feat); ret_lst <- regexec(rex_str, txt_vctr); ret_lst <- regmatches(txt_vctr, ret_lst); ret_vctr <- sapply(1:length(ret_lst), function(pos_ix) ifelse(length(ret_lst[[pos_ix]]) > 0, ret_lst[[pos_ix]], "")); print(ret_vctr <- ret_vctr[ret_vctr != ""])
if (nrow(featsimp_df) > 5) {
warning("Limiting important feature scatter plots to 5 out of ", nrow(featsimp_df))
featsimp_df <- head(featsimp_df, 5)
}
# if (!all(is.na(featsimp_df$feat.interact)))
# stop("not implemented yet")
rsp_var_out <- paste0(glb_rsp_var_out, mdl_id)
for (var in featsimp_df$feat) {
plot_df <- melt(obs_df, id.vars=var,
measure.vars=c(glb_rsp_var, rsp_var_out))
# if (var == "<feat_name>") print(myplot_scatter(plot_df, var, "value",
# facet_colcol_name="variable") +
# geom_vline(xintercept=<divider_val>, linetype="dotted")) else
print(myplot_scatter(plot_df, var, "value", colorcol_name="variable",
facet_colcol_name="variable", jitter=TRUE) +
guides(color=FALSE))
}
if (glb_is_regression) {
if (nrow(featsimp_df) == 0)
warning("No important features in glb_fin_mdl") else
print(myplot_prediction_regression(df=obs_df,
feat_x=ifelse(nrow(featsimp_df) > 1, featsimp_df$feat[2],
".rownames"),
feat_y=featsimp_df$feat[1],
rsp_var=glb_rsp_var, rsp_var_out=rsp_var_out,
id_vars=glb_id_var)
# + facet_wrap(reformulate(featsimp_df$feat[2])) # if [1 or 2] is a factor
# + geom_point(aes_string(color="<col_name>.fctr")) # to color the plot
)
}
if (glb_is_classification) {
if (nrow(featsimp_df) == 0)
warning("No features in selected model are statistically important")
else print(myplot_prediction_classification(df=obs_df,
feat_x=ifelse(nrow(featsimp_df) > 1, featsimp_df$feat[2],
".rownames"),
feat_y=featsimp_df$feat[1],
rsp_var=glb_rsp_var,
rsp_var_out=rsp_var_out,
id_vars=glb_id_var,
prob_threshold=prob_threshold)
# + geom_hline(yintercept=<divider_val>, linetype = "dotted")
)
}
}
if (glb_is_classification && glb_is_binomial)
glb_analytics_diag_plots(obs_df=glb_OOBobs_df, mdl_id=glb_sel_mdl_id,
prob_threshold=glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"opt.prob.threshold.OOB"]) else
glb_analytics_diag_plots(obs_df=glb_OOBobs_df, mdl_id=glb_sel_mdl_id) 
## [1] "Min/Max Boundaries: "
## .rownames EmploymentStatus.fctr
## 232 83771 Retired
## 933 80246 Not.in.Labor.Force
## 4110 111997 Disabled
## 5360 55065 Disabled
## 12602 41421 Disabled
## 18898 1371 Disabled
## 26299 20313 Not.in.Labor.Force
## 13635 92281 Unemployed
## 26264 61504 Unemployed
## 27111 57730 Disabled
## 31868 66224 Disabled
## 41907 49837 Disabled
## 43708 66821 Disabled
## 12806 39311 Unemployed
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## 232 Retired
## 933 Not.in.Labor.Force
## 4110 Retired
## 5360 Retired
## 12602 Retired
## 18898 Retired
## 26299 Retired
## 13635 Not.in.Labor.Force
## 26264 Not.in.Labor.Force
## 27111 Not.in.Labor.Force
## 31868 Not.in.Labor.Force
## 41907 Not.in.Labor.Force
## 43708 Not.in.Labor.Force
## 12806 Employed
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## 232 0.043811929
## 933 0.015333504
## 4110 0.043811929
## 5360 0.043811929
## 12602 0.043811929
## 18898 0.043811929
## 26299 0.043811929
## 13635 0.015333504
## 26264 0.015333504
## 27111 0.015333504
## 31868 0.015333504
## 41907 0.015333504
## 43708 0.015333504
## 12806 0.002604167
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## 232 TRUE
## 933 TRUE
## 4110 FALSE
## 5360 FALSE
## 12602 FALSE
## 18898 FALSE
## 26299 FALSE
## 13635 FALSE
## 26264 FALSE
## 27111 FALSE
## 31868 FALSE
## 41907 FALSE
## 43708 FALSE
## 12806 FALSE
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.error .label
## 232 0.0000000 83771
## 933 0.0000000 80246
## 4110 0.9561881 111997
## 5360 0.9561881 55065
## 12602 0.9561881 41421
## 18898 0.9561881 1371
## 26299 0.9561881 20313
## 13635 0.9846665 92281
## 26264 0.9846665 61504
## 27111 0.9846665 57730
## 31868 0.9846665 66224
## 41907 0.9846665 49837
## 43708 0.9846665 66821
## 12806 0.9973958 39311
## [1] "Inaccurate: "
## .rownames EmploymentStatus.fctr
## 1751 1258 Retired
## 2857 1287 Retired
## 4163 110947 Retired
## 5504 56944 Retired
## 5947 56276 Not.in.Labor.Force
## 7059 286 Not.in.Labor.Force
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## 1751 Disabled
## 2857 Disabled
## 4163 Disabled
## 5504 Disabled
## 5947 Disabled
## 7059 Disabled
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## 1751 0.4559859
## 2857 0.4559859
## 4163 0.4559859
## 5504 0.4559859
## 5947 0.4559859
## 7059 0.4559859
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## 1751 FALSE
## 2857 FALSE
## 4163 FALSE
## 5504 FALSE
## 5947 FALSE
## 7059 FALSE
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.error
## 1751 0.5440141
## 2857 0.5440141
## 4163 0.5440141
## 5504 0.5440141
## 5947 0.5440141
## 7059 0.5440141
## .rownames EmploymentStatus.fctr
## 48873 107691 Not.in.Labor.Force
## 90321 126562 Disabled
## 32989 100297 Unemployed
## 7390 436 Unemployed
## 63675 96452 Unemployed
## 71179 47259 Unemployed
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## 48873 Disabled
## 90321 Retired
## 32989 Not.in.Labor.Force
## 7390 Employed
## 63675 Employed
## 71179 Employed
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## 48873 0.4457953394
## 90321 0.0438119293
## 32989 0.0153335037
## 7390 0.0012799415
## 63675 0.0007094714
## 71179 0.0004582951
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## 48873 FALSE
## 90321 FALSE
## 32989 FALSE
## 7390 FALSE
## 63675 FALSE
## 71179 FALSE
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.error
## 48873 0.5542047
## 90321 0.9561881
## 32989 0.9846665
## 7390 0.9987201
## 63675 0.9992905
## 71179 0.9995417
## .rownames EmploymentStatus.fctr
## 129200 73943 Unemployed
## 129586 21684 Unemployed
## 130229 58290 Unemployed
## 130736 20447 Unemployed
## 130934 33835 Unemployed
## 131084 40385 Disabled
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## 129200 Employed
## 129586 Employed
## 130229 Employed
## 130736 Employed
## 130934 Employed
## 131084 Employed
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## 129200 0
## 129586 0
## 130229 0
## 130736 0
## 130934 0
## 131084 0
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## 129200 FALSE
## 129586 FALSE
## 130229 FALSE
## 130736 FALSE
## 130934 FALSE
## 131084 FALSE
## EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.error
## 129200 1
## 129586 1
## 130229 1
## 130736 1
## 130934 1
## 131084 1
# gather predictions from models better than MFO.*
#mdl_id <- "Conditional.X.rf"
#mdl_id <- "Conditional.X.cp.0.rpart"
#mdl_id <- "Conditional.X.rpart"
# glb_OOBobs_df <- glb_get_predictions(df=glb_OOBobs_df, mdl_id,
# glb_rsp_var_out)
# print(t(confusionMatrix(glb_OOBobs_df[, paste0(glb_rsp_var_out, mdl_id)],
# glb_OOBobs_df[, glb_rsp_var])$table))
FN_OOB_ids <- c(4721, 4020, 693, 92)
print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
grep(glb_rsp_var, names(glb_OOBobs_df), value=TRUE)])## [1] EmploymentStatus.fctr
## [2] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## [3] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## [4] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## <0 rows> (or 0-length row.names)print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
glb_feats_df$id[1:5]])## [1] Age Sex.fctr Married.my.fctr Region.fctr
## [5] .rnorm
## <0 rows> (or 0-length row.names)print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
glb_txt_vars])## data frame with 0 columns and 0 rowswrite.csv(glb_OOBobs_df[, c(glb_id_var,
grep(glb_rsp_var, names(glb_OOBobs_df), fixed=TRUE, value=TRUE))],
paste0(gsub(".", "_", paste0(glb_out_pfx, glb_sel_mdl_id), fixed=TRUE),
"_OOBobs.csv"), row.names=FALSE)
# print(glb_allobs_df[glb_allobs_df$UniqueID %in% FN_OOB_ids,
# glb_txt_vars])
# dsp_tbl(Headline.contains="[Ee]bola")
# sum(sel_obs(Headline.contains="[Ee]bola"))
# ftable(xtabs(Popular ~ NewsDesk.fctr, data=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,]))
# xtabs(NewsDesk ~ Popular, #Popular ~ NewsDesk.fctr,
# data=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,],
# exclude=NULL)
# print(mycreate_xtab_df(df=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,], c("Popular", "NewsDesk", "SectionName", "SubsectionName")))
# print(mycreate_tbl_df(df=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,], c("Popular", "NewsDesk", "SectionName", "SubsectionName")))
# print(mycreate_tbl_df(df=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,], c("Popular")))
# print(mycreate_tbl_df(df=glb_allobs_df[sel_obs(Headline.contains="[Ee]bola") ,],
# tbl_col_names=c("Popular", "NewsDesk")))
glb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.models", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 12 fit.models 7 2 244.919 268.677 23.758
## 13 fit.models 7 3 268.677 NA NAprint(setdiff(names(glb_trnobs_df), names(glb_allobs_df)))## character(0)print(setdiff(names(glb_fitobs_df), names(glb_allobs_df)))## character(0)print(setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))## [1] "EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart"
## [2] "EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob"
## [3] "EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate"for (col in setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))
# Merge or cbind ?
glb_allobs_df[glb_allobs_df$.lcn == "OOB", col] <- glb_OOBobs_df[, col]
print(setdiff(names(glb_newobs_df), names(glb_allobs_df)))## character(0)if (glb_save_envir)
save(glb_feats_df,
glb_allobs_df, #glb_trnobs_df, glb_fitobs_df, glb_OOBobs_df, glb_newobs_df,
glb_models_df, dsp_models_df, glb_models_lst, glb_sel_mdl, glb_sel_mdl_id,
glb_model_type,
file=paste0(glb_out_pfx, "selmdl_dsk.RData"))
#load(paste0(glb_out_pfx, "selmdl_dsk.RData"))
rm(ret_lst)## Warning in rm(ret_lst): object 'ret_lst' not foundreplay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"model.selected")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
## 2.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction data.new.prediction firing: model.selected
## 3.0000 3 0 2 1 0
glb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.data.training", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 13 fit.models 7 3 268.677 281.596 12.92
## 14 fit.data.training 8 0 281.597 NA NAStep 8.0: fit data training
#load(paste0(glb_inp_pfx, "dsk.RData"))
# To create specific models
# glb_fin_mdl_id <- NULL; glb_fin_mdl <- NULL;
# glb_sel_mdl_id <- "Conditional.X.cp.0.rpart";
# glb_sel_mdl <- glb_models_lst[[glb_sel_mdl_id]]; print(glb_sel_mdl)
if (!is.null(glb_fin_mdl_id) && (glb_fin_mdl_id %in% names(glb_models_lst))) {
warning("Final model same as user selected model")
glb_fin_mdl <- glb_sel_mdl
} else {
# print(mdl_feats_df <- myextract_mdl_feats(sel_mdl=glb_sel_mdl,
# entity_df=glb_fitobs_df))
if ((model_method <- glb_sel_mdl$method) == "custom")
# get actual method from the model_id
model_method <- tail(unlist(strsplit(glb_sel_mdl_id, "[.]")), 1)
tune_finmdl_df <- NULL
if (nrow(glb_sel_mdl$bestTune) > 0) {
for (param in names(glb_sel_mdl$bestTune)) {
#print(sprintf("param: %s", param))
if (glb_sel_mdl$bestTune[1, param] != "none")
tune_finmdl_df <- rbind(tune_finmdl_df,
data.frame(parameter=param,
min=glb_sel_mdl$bestTune[1, param],
max=glb_sel_mdl$bestTune[1, param],
by=1)) # by val does not matter
}
}
# Sync with parameters in mydsutils.R
require(gdata)
ret_lst <- myfit_mdl(model_id="Final", model_method=model_method,
indep_vars_vctr=trim(unlist(strsplit(glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"feats"], "[,]"))),
model_type=glb_model_type,
rsp_var=glb_rsp_var, rsp_var_out=glb_rsp_var_out,
fit_df=glb_trnobs_df, OOB_df=NULL,
n_cv_folds=glb_n_cv_folds, tune_models_df=tune_finmdl_df,
# Automate from here
# Issues if glb_sel_mdl$method == "rf" b/c trainControl is "oob"; not "cv"
model_loss_mtrx=glb_model_metric_terms,
model_summaryFunction=glb_sel_mdl$control$summaryFunction,
model_metric=glb_sel_mdl$metric,
model_metric_maximize=glb_sel_mdl$maximize)
glb_fin_mdl <- glb_models_lst[[length(glb_models_lst)]]
glb_fin_mdl_id <- glb_models_df[length(glb_models_lst), "model_id"]
}## Loading required package: gdata
## gdata: read.xls support for 'XLS' (Excel 97-2004) files ENABLED.
##
## gdata: read.xls support for 'XLSX' (Excel 2007+) files ENABLED.
##
## Attaching package: 'gdata'
##
## The following objects are masked from 'package:dplyr':
##
## combine, first, last
##
## The following object is masked from 'package:stats':
##
## nobs
##
## The following object is masked from 'package:utils':
##
## object.size## [1] "fitting model: Final.rpart"
## [1] " indep_vars: Industry.my.fctr, Age"
## Aggregating results
## Fitting final model on full training set## Call:
## rpart(formula = .outcome ~ ., control = list(minsplit = 20, minbucket = 7,
## cp = 0, maxcompete = 4, maxsurrogate = 5, usesurrogate = 2,
## surrogatestyle = 0, maxdepth = 30, xval = 0))
## n= 105513
##
## CP nsplit rel error
## 1 0.2499771585 0 1.0000000
## 2 0.0888533577 1 0.7500228
## 3 0.0227272727 2 0.6611695
## 4 0.0167884879 3 0.6384422
## 5 0.0135449977 4 0.6216537
## 6 0.0127912289 5 0.6081087
## 7 0.0102558246 6 0.5953175
## 8 0.0089107152 7 0.5850617
## 9 0.0080630425 23 0.3582001
## 10 0.0076975788 24 0.3501370
## 11 0.0073777981 25 0.3424395
## 12 0.0071950662 26 0.3350617
## 13 0.0066240292 27 0.3278666
## 14 0.0052992234 28 0.3212426
## 15 0.0046596619 29 0.3159434
## 16 0.0046368205 30 0.3112837
## 17 0.0045911375 31 0.3066469
## 18 0.0037460027 32 0.3020557
## 19 0.0026496117 33 0.2983097
## 20 0.0020100503 34 0.2956601
## 21 0.0019186843 35 0.2936501
## 22 0.0016902695 36 0.2917314
## 23 0.0015303792 37 0.2900411
## 24 0.0013933303 38 0.2885107
## 25 0.0010963910 39 0.2871174
## 26 0.0009821836 40 0.2860210
## 27 0.0007994518 41 0.2850388
## 28 0.0007766103 42 0.2842394
## 29 0.0007537688 43 0.2834628
## 30 0.0002512563 44 0.2827090
## 31 0.0002055733 45 0.2824577
## 32 0.0001713111 47 0.2820466
## 33 0.0001598904 49 0.2817040
## 34 0.0001370489 50 0.2815441
## 35 0.0000000000 51 0.2814070
##
## Variable importance
## Age
## 42
## Industry.my.fctrOther services
## 6
## Industry.my.fctrEducational and health services
## 6
## Industry.my.fctrLeisure and hospitality
## 5
## Industry.my.fctrTransportation and utilities
## 5
## Industry.my.fctrTrade
## 5
## Industry.my.fctrConstruction
## 4
## Industry.my.fctrPublic administration
## 4
## Industry.my.fctrFinancial
## 4
## Industry.my.fctrProfessional and business services
## 4
## Industry.my.fctrManufacturing
## 4
## Industry.my.fctrAgriculture, forestry, fishing, and hunting
## 4
## Industry.my.fctrInformation
## 3
## Industry.my.fctrMining
## 2
##
## Node number 1: 105513 observations, complexity param=0.2499772
## predicted class=Employed expected loss=0.4149252 P(node) =1
## class counts: 5712 61733 15246 18619 4203
## probabilities: 0.054 0.585 0.144 0.176 0.040
## left son=2 (83904 obs) right son=3 (21609 obs)
## Primary splits:
## Age < 63.5 to the left, improve=12147.510, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve= 3188.841, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 1623.725, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 1349.429, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 1254.102, (0 missing)
##
## Node number 2: 83904 observations, complexity param=0.08885336
## predicted class=Employed expected loss=0.3176249 P(node) =0.7952006
## class counts: 4727 57254 14741 3196 3986
## probabilities: 0.056 0.682 0.176 0.038 0.048
## left son=4 (75489 obs) right son=5 (8415 obs)
## Primary splits:
## Age < 19.5 to the right, improve=4446.9060, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1719.4460, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 826.8062, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 684.5403, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 663.4894, (0 missing)
##
## Node number 3: 21609 observations, complexity param=0.02272727
## predicted class=Retired expected loss=0.2862696 P(node) =0.2047994
## class counts: 985 4479 505 15423 217
## probabilities: 0.046 0.207 0.023 0.714 0.010
## left son=6 (1104 obs) right son=7 (20505 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1157.8420, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 677.5990, (0 missing)
## Age < 70.5 to the left, improve= 657.3056, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 635.9218, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 399.3432, (0 missing)
##
## Node number 4: 75489 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.2676681 P(node) =0.7154474
## class counts: 4622 55283 8880 3188 3516
## probabilities: 0.061 0.732 0.118 0.042 0.047
## left son=8 (13669 obs) right son=9 (61820 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=1092.0010, (0 missing)
## Age < 57.5 to the left, improve= 564.9899, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve= 477.2123, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 418.9047, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve= 405.9581, (0 missing)
##
## Node number 5: 8415 observations, complexity param=0.01678849
## predicted class=Not.in.Labor.Force expected loss=0.3035056 P(node) =0.07975321
## class counts: 105 1971 5861 8 470
## probabilities: 0.012 0.234 0.696 0.001 0.056
## left son=10 (930 obs) right son=11 (7485 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=821.24650, (0 missing)
## Industry.my.fctrTrade < 0.5 to the right, improve=483.79540, (0 missing)
## Age < 17.5 to the right, improve=315.44200, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=162.07580, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 87.34646, (0 missing)
##
## Node number 6: 1104 observations
## predicted class=Employed expected loss=0.07065217 P(node) =0.01046317
## class counts: 0 1026 0 31 47
## probabilities: 0.000 0.929 0.000 0.028 0.043
##
## Node number 7: 20505 observations, complexity param=0.013545
## predicted class=Retired expected loss=0.2493538 P(node) =0.1943362
## class counts: 985 3453 505 15392 170
## probabilities: 0.048 0.168 0.025 0.751 0.008
## left son=14 (647 obs) right son=15 (19858 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=752.9797, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=707.5644, (0 missing)
## Age < 69.5 to the left, improve=469.6243, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=443.2823, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=404.4244, (0 missing)
##
## Node number 8: 13669 observations
## predicted class=Employed expected loss=0.05004024 P(node) =0.129548
## class counts: 13 12985 65 18 588
## probabilities: 0.001 0.950 0.005 0.001 0.043
##
## Node number 9: 61820 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.3157878 P(node) =0.5858994
## class counts: 4609 42298 8815 3170 2928
## probabilities: 0.075 0.684 0.143 0.051 0.047
## left son=18 (7743 obs) right son=19 (54077 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=751.9581, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=655.3610, (0 missing)
## Age < 57.5 to the left, improve=651.8145, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=629.3461, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=439.6490, (0 missing)
##
## Node number 10: 930 observations
## predicted class=Employed expected loss=0.1462366 P(node) =0.00881408
## class counts: 0 794 59 0 77
## probabilities: 0.000 0.854 0.063 0.000 0.083
##
## Node number 11: 7485 observations, complexity param=0.01025582
## predicted class=Not.in.Labor.Force expected loss=0.2248497 P(node) =0.07093913
## class counts: 105 1177 5802 8 393
## probabilities: 0.014 0.157 0.775 0.001 0.053
## left son=22 (543 obs) right son=23 (6942 obs)
## Primary splits:
## Industry.my.fctrTrade < 0.5 to the right, improve=612.76670, (0 missing)
## Age < 17.5 to the right, improve=239.19990, (0 missing)
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=210.34410, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=112.11510, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve= 95.73304, (0 missing)
##
## Node number 14: 647 observations
## predicted class=Employed expected loss=0.05873261 P(node) =0.006131946
## class counts: 3 609 1 16 18
## probabilities: 0.005 0.941 0.002 0.025 0.028
##
## Node number 15: 19858 observations, complexity param=0.01279123
## predicted class=Retired expected loss=0.2257025 P(node) =0.1882043
## class counts: 982 2844 504 15376 152
## probabilities: 0.049 0.143 0.025 0.774 0.008
## left son=30 (624 obs) right son=31 (19234 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=755.7690, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=472.8057, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=431.4262, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=419.6137, (0 missing)
## Age < 69.5 to the left, improve=392.8600, (0 missing)
##
## Node number 18: 7743 observations
## predicted class=Employed expected loss=0.07464807 P(node) =0.07338432
## class counts: 10 7165 50 6 512
## probabilities: 0.001 0.925 0.006 0.001 0.066
##
## Node number 19: 54077 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.3503153 P(node) =0.512515
## class counts: 4599 35133 8765 3164 2416
## probabilities: 0.085 0.650 0.162 0.059 0.045
## left son=38 (6767 obs) right son=39 (47310 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=870.1717, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=830.7705, (0 missing)
## Age < 57.5 to the left, improve=692.9737, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=575.1856, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=565.4602, (0 missing)
##
## Node number 22: 543 observations
## predicted class=Employed expected loss=0.1362799 P(node) =0.005146285
## class counts: 0 469 20 0 54
## probabilities: 0.000 0.864 0.037 0.000 0.099
##
## Node number 23: 6942 observations, complexity param=0.003746003
## predicted class=Not.in.Labor.Force expected loss=0.1670988 P(node) =0.06579284
## class counts: 105 708 5782 8 339
## probabilities: 0.015 0.102 0.833 0.001 0.049
## left son=46 (244 obs) right son=47 (6698 obs)
## Primary splits:
## Industry.my.fctrEducational and health services < 0.5 to the right, improve=249.47940, (0 missing)
## Age < 17.5 to the right, improve=136.16640, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=132.10790, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=114.47680, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 99.72118, (0 missing)
##
## Node number 30: 624 observations
## predicted class=Employed expected loss=0.08173077 P(node) =0.005913963
## class counts: 1 573 4 13 33
## probabilities: 0.002 0.918 0.006 0.021 0.053
##
## Node number 31: 19234 observations, complexity param=0.008063042
## predicted class=Retired expected loss=0.2012582 P(node) =0.1822903
## class counts: 981 2271 500 15363 119
## probabilities: 0.051 0.118 0.026 0.799 0.006
## left son=62 (383 obs) right son=63 (18851 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=503.8076, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=459.7628, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=447.4862, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=411.0300, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=330.4545, (0 missing)
##
## Node number 38: 6767 observations
## predicted class=Employed expected loss=0.07285355 P(node) =0.06413428
## class counts: 7 6274 35 3 448
## probabilities: 0.001 0.927 0.005 0.000 0.066
##
## Node number 39: 47310 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.3900021 P(node) =0.4483808
## class counts: 4592 28859 8730 3161 1968
## probabilities: 0.097 0.610 0.185 0.067 0.042
## left son=78 (6319 obs) right son=79 (40991 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=1101.9840, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve= 772.3607, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 731.6832, (0 missing)
## Age < 57.5 to the left, improve= 724.2414, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 634.3962, (0 missing)
##
## Node number 46: 244 observations
## predicted class=Employed expected loss=0.2090164 P(node) =0.002312511
## class counts: 0 193 29 0 22
## probabilities: 0.000 0.791 0.119 0.000 0.090
##
## Node number 47: 6698 observations, complexity param=0.00201005
## predicted class=Not.in.Labor.Force expected loss=0.1410869 P(node) =0.06348033
## class counts: 105 515 5753 8 317
## probabilities: 0.016 0.077 0.859 0.001 0.047
## left son=94 (124 obs) right son=95 (6574 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=141.70320, (0 missing)
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=123.52010, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=106.74930, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=101.13560, (0 missing)
## Age < 17.5 to the right, improve= 93.61962, (0 missing)
##
## Node number 62: 383 observations
## predicted class=Employed expected loss=0.06788512 P(node) =0.003629884
## class counts: 0 357 2 4 20
## probabilities: 0.000 0.932 0.005 0.010 0.052
##
## Node number 63: 18851 observations, complexity param=0.007377798
## predicted class=Retired expected loss=0.1852422 P(node) =0.1786604
## class counts: 981 1914 498 15359 99
## probabilities: 0.052 0.102 0.026 0.815 0.005
## left son=126 (352 obs) right son=127 (18499 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=478.8641, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=466.2788, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=428.3619, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=343.8240, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=300.2703, (0 missing)
##
## Node number 78: 6319 observations
## predicted class=Employed expected loss=0.06678272 P(node) =0.05988835
## class counts: 8 5897 21 10 383
## probabilities: 0.001 0.933 0.003 0.002 0.061
##
## Node number 79: 40991 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.4398283 P(node) =0.3884924
## class counts: 4584 22962 8709 3151 1585
## probabilities: 0.112 0.560 0.212 0.077 0.039
## left son=158 (5109 obs) right son=159 (35882 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=1064.4410, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 973.0519, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve= 864.2965, (0 missing)
## Age < 55.5 to the left, improve= 797.2025, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 719.5783, (0 missing)
##
## Node number 94: 124 observations
## predicted class=Employed expected loss=0.2096774 P(node) =0.001175211
## class counts: 0 98 10 0 16
## probabilities: 0.000 0.790 0.081 0.000 0.129
##
## Node number 95: 6574 observations, complexity param=0.00169027
## predicted class=Not.in.Labor.Force expected loss=0.1264071 P(node) =0.06230512
## class counts: 105 417 5743 8 301
## probabilities: 0.016 0.063 0.874 0.001 0.046
## left son=190 (128 obs) right son=191 (6446 obs)
## Primary splits:
## Industry.my.fctrProfessional and business services < 0.5 to the right, improve=128.68200, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=110.73290, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=104.99030, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 81.24365, (0 missing)
## Age < 17.5 to the right, improve= 75.73573, (0 missing)
##
## Node number 126: 352 observations
## predicted class=Employed expected loss=0.05965909 P(node) =0.003336082
## class counts: 0 331 1 8 12
## probabilities: 0.000 0.940 0.003 0.023 0.034
##
## Node number 127: 18499 observations, complexity param=0.007195066
## predicted class=Retired expected loss=0.1701714 P(node) =0.1753244
## class counts: 981 1583 497 15351 87
## probabilities: 0.053 0.086 0.027 0.830 0.005
## left son=254 (350 obs) right son=255 (18149 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=484.5522, (0 missing)
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=445.2126, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=356.8226, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=311.9446, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=310.0414, (0 missing)
##
## Node number 158: 5109 observations
## predicted class=Employed expected loss=0.09062439 P(node) =0.04842057
## class counts: 3 4646 54 6 400
## probabilities: 0.001 0.909 0.011 0.001 0.078
##
## Node number 159: 35882 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.4895491 P(node) =0.3400718
## class counts: 4581 18316 8655 3145 1185
## probabilities: 0.128 0.510 0.241 0.088 0.033
## left son=318 (3949 obs) right son=319 (31933 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=1257.5160, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=1139.3140, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve= 923.8724, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 883.1958, (0 missing)
## Age < 54.5 to the left, improve= 794.6089, (0 missing)
##
## Node number 190: 128 observations
## predicted class=Employed expected loss=0.296875 P(node) =0.001213121
## class counts: 0 90 16 0 22
## probabilities: 0.000 0.703 0.125 0.000 0.172
##
## Node number 191: 6446 observations, complexity param=0.001530379
## predicted class=Not.in.Labor.Force expected loss=0.111542 P(node) =0.061092
## class counts: 105 327 5727 8 279
## probabilities: 0.016 0.051 0.888 0.001 0.043
## left son=382 (89 obs) right son=383 (6357 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=114.72730, (0 missing)
## Industry.my.fctrManufacturing < 0.5 to the right, improve=108.88810, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 84.40884, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 72.69765, (0 missing)
## Age < 17.5 to the right, improve= 57.89279, (0 missing)
##
## Node number 254: 350 observations
## predicted class=Employed expected loss=0.07714286 P(node) =0.003317127
## class counts: 0 323 0 8 19
## probabilities: 0.000 0.923 0.000 0.023 0.054
##
## Node number 255: 18149 observations, complexity param=0.006624029
## predicted class=Retired expected loss=0.1546091 P(node) =0.1720072
## class counts: 981 1260 497 15343 68
## probabilities: 0.054 0.069 0.027 0.845 0.004
## left son=510 (325 obs) right son=511 (17824 obs)
## Primary splits:
## Industry.my.fctrLeisure and hospitality < 0.5 to the right, improve=462.7893, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=370.3579, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=324.1177, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=322.1780, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=319.2237, (0 missing)
##
## Node number 318: 3949 observations
## predicted class=Employed expected loss=0.03925044 P(node) =0.03742667
## class counts: 0 3794 12 2 141
## probabilities: 0.000 0.961 0.003 0.001 0.036
##
## Node number 319: 31933 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.5452353 P(node) =0.3026452
## class counts: 4581 14522 8643 3143 1044
## probabilities: 0.143 0.455 0.271 0.098 0.033
## left son=638 (4073 obs) right son=639 (27860 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=1482.8070, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1177.6570, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=1133.7060, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve= 994.4203, (0 missing)
## Age < 53.5 to the left, improve= 854.2872, (0 missing)
##
## Node number 382: 89 observations
## predicted class=Employed expected loss=0.1797753 P(node) =0.000843498
## class counts: 0 73 6 0 10
## probabilities: 0.000 0.820 0.067 0.000 0.112
##
## Node number 383: 6357 observations, complexity param=0.00139333
## predicted class=Not.in.Labor.Force expected loss=0.1000472 P(node) =0.0602485
## class counts: 105 254 5721 8 269
## probabilities: 0.017 0.040 0.900 0.001 0.042
## left son=766 (89 obs) right son=767 (6268 obs)
## Primary splits:
## Industry.my.fctrManufacturing < 0.5 to the right, improve=112.03640, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 86.98183, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve= 74.58660, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve= 59.02914, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 57.30520, (0 missing)
##
## Node number 510: 325 observations
## predicted class=Employed expected loss=0.08615385 P(node) =0.003080189
## class counts: 1 297 0 7 20
## probabilities: 0.003 0.914 0.000 0.022 0.062
##
## Node number 511: 17824 observations, complexity param=0.005299223
## predicted class=Retired expected loss=0.1395871 P(node) =0.1689271
## class counts: 980 963 497 15336 48
## probabilities: 0.055 0.054 0.028 0.860 0.003
## left son=1022 (243 obs) right son=1023 (17581 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=383.5801, (0 missing)
## Industry.my.fctrConstruction < 0.5 to the right, improve=336.0193, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=334.0430, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=330.8698, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=137.2679, (0 missing)
##
## Node number 638: 4073 observations
## predicted class=Employed expected loss=0.08102136 P(node) =0.03860188
## class counts: 2 3743 19 9 300
## probabilities: 0.000 0.919 0.005 0.002 0.074
##
## Node number 639: 27860 observations, complexity param=0.008910715
## predicted class=Employed expected loss=0.6131012 P(node) =0.2640433
## class counts: 4579 10779 8624 3134 744
## probabilities: 0.164 0.387 0.310 0.112 0.027
## left son=1278 (2949 obs) right son=1279 (24911 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=1543.6710, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=1497.3620, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=1317.7560, (0 missing)
## Age < 50.5 to the left, improve= 939.5404, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 531.7269, (0 missing)
##
## Node number 766: 89 observations
## predicted class=Employed expected loss=0.2696629 P(node) =0.000843498
## class counts: 0 65 4 0 20
## probabilities: 0.000 0.730 0.045 0.000 0.225
##
## Node number 767: 6268 observations, complexity param=0.001096391
## predicted class=Not.in.Labor.Force expected loss=0.08790683 P(node) =0.059405
## class counts: 105 189 5717 8 249
## probabilities: 0.017 0.030 0.912 0.001 0.040
## left son=1534 (76 obs) right son=1535 (6192 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=89.59828, (0 missing)
## Industry.my.fctrFinancial < 0.5 to the right, improve=76.47611, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=60.57117, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=58.68684, (0 missing)
## Age < 17.5 to the left, improve=26.06129, (0 missing)
##
## Node number 1022: 243 observations
## predicted class=Employed expected loss=0.03292181 P(node) =0.002303034
## class counts: 0 235 0 3 5
## probabilities: 0.000 0.967 0.000 0.012 0.021
##
## Node number 1023: 17581 observations, complexity param=0.004659662
## predicted class=Retired expected loss=0.1278653 P(node) =0.166624
## class counts: 980 728 497 15333 43
## probabilities: 0.056 0.041 0.028 0.872 0.002
## left son=2046 (225 obs) right son=2047 (17356 obs)
## Primary splits:
## Industry.my.fctrConstruction < 0.5 to the right, improve=345.6840, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=343.6840, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=340.3347, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=141.2113, (0 missing)
## Age < 69.5 to the left, improve=115.1508, (0 missing)
##
## Node number 1278: 2949 observations
## predicted class=Employed expected loss=0.0342489 P(node) =0.02794916
## class counts: 4 2848 7 8 82
## probabilities: 0.001 0.966 0.002 0.003 0.028
##
## Node number 1279: 24911 observations, complexity param=0.008910715
## predicted class=Not.in.Labor.Force expected loss=0.6540886 P(node) =0.2360941
## class counts: 4575 7931 8617 3126 662
## probabilities: 0.184 0.318 0.346 0.125 0.027
## left son=2558 (2995 obs) right son=2559 (21916 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=1909.4590, (0 missing)
## Industry.my.fctrOther services < 0.5 to the right, improve=1683.7930, (0 missing)
## Age < 50.5 to the left, improve=1038.2920, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 675.5507, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 564.8951, (0 missing)
##
## Node number 1534: 76 observations
## predicted class=Employed expected loss=0.2631579 P(node) =0.0007202904
## class counts: 0 56 8 0 12
## probabilities: 0.000 0.737 0.105 0.000 0.158
##
## Node number 1535: 6192 observations, complexity param=0.0009821836
## predicted class=Not.in.Labor.Force expected loss=0.07800388 P(node) =0.05868471
## class counts: 105 133 5709 8 237
## probabilities: 0.017 0.021 0.922 0.001 0.038
## left son=3070 (48 obs) right son=3071 (6144 obs)
## Primary splits:
## Industry.my.fctrFinancial < 0.5 to the right, improve=78.09325, (0 missing)
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=61.88774, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=59.87133, (0 missing)
## Age < 17.5 to the left, improve=21.60476, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=20.63575, (0 missing)
##
## Node number 2046: 225 observations
## predicted class=Employed expected loss=0.08 P(node) =0.002132439
## class counts: 0 207 1 3 14
## probabilities: 0.000 0.920 0.004 0.013 0.062
##
## Node number 2047: 17356 observations, complexity param=0.00463682
## predicted class=Retired expected loss=0.116732 P(node) =0.1644916
## class counts: 980 521 496 15330 29
## probabilities: 0.056 0.030 0.029 0.883 0.002
## left son=4094 (225 obs) right son=4095 (17131 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=352.73740, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=349.21860, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=144.91390, (0 missing)
## Age < 69.5 to the left, improve= 93.27941, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 36.16099, (0 missing)
##
## Node number 2558: 2995 observations
## predicted class=Employed expected loss=0.05409015 P(node) =0.02838513
## class counts: 3 2833 8 4 147
## probabilities: 0.001 0.946 0.003 0.001 0.049
##
## Node number 2559: 21916 observations, complexity param=0.008910715
## predicted class=Not.in.Labor.Force expected loss=0.607182 P(node) =0.207709
## class counts: 4572 5098 8609 3122 515
## probabilities: 0.209 0.233 0.393 0.142 0.023
## left son=5118 (2748 obs) right son=5119 (19168 obs)
## Primary splits:
## Industry.my.fctrOther services < 0.5 to the right, improve=2220.5360, (0 missing)
## Age < 50.5 to the left, improve=1171.3010, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve= 883.7817, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 735.8725, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 388.6810, (0 missing)
##
## Node number 3070: 48 observations
## predicted class=Employed expected loss=0.08333333 P(node) =0.0004549202
## class counts: 0 44 1 0 3
## probabilities: 0.000 0.917 0.021 0.000 0.063
##
## Node number 3071: 6144 observations, complexity param=0.0007766103
## predicted class=Not.in.Labor.Force expected loss=0.07096354 P(node) =0.05822979
## class counts: 105 89 5708 8 234
## probabilities: 0.017 0.014 0.929 0.001 0.038
## left son=6142 (40 obs) right son=6143 (6104 obs)
## Primary splits:
## Industry.my.fctrTransportation and utilities < 0.5 to the right, improve=62.88517, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=60.77505, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=21.02117, (0 missing)
## Industry.my.fctrMining < 0.5 to the left, improve=15.99500, (0 missing)
## Age < 17.5 to the left, improve=14.80758, (0 missing)
##
## Node number 4094: 225 observations
## predicted class=Employed expected loss=0.07555556 P(node) =0.002132439
## class counts: 0 208 1 5 11
## probabilities: 0.000 0.924 0.004 0.022 0.049
##
## Node number 4095: 17131 observations, complexity param=0.004591138
## predicted class=Retired expected loss=0.1054229 P(node) =0.1623591
## class counts: 980 313 495 15325 18
## probabilities: 0.057 0.018 0.029 0.895 0.001
## left son=8190 (219 obs) right son=8191 (16912 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=358.43180, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=148.75290, (0 missing)
## Age < 68.5 to the left, improve= 73.21838, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 37.16997, (0 missing)
##
## Node number 5118: 2748 observations
## predicted class=Employed expected loss=0.06550218 P(node) =0.02604418
## class counts: 2 2568 17 1 160
## probabilities: 0.001 0.934 0.006 0.000 0.058
##
## Node number 5119: 19168 observations, complexity param=0.008910715
## predicted class=Not.in.Labor.Force expected loss=0.5517529 P(node) =0.1816648
## class counts: 4570 2530 8592 3121 355
## probabilities: 0.238 0.132 0.448 0.163 0.019
## left son=10238 (7944 obs) right son=10239 (11224 obs)
## Primary splits:
## Age < 49.5 to the right, improve=1339.92900, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=1167.34600, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve= 967.76410, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 506.78100, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 30.53093, (0 missing)
##
## Node number 6142: 40 observations
## predicted class=Employed expected loss=0.125 P(node) =0.0003791002
## class counts: 0 35 1 0 4
## probabilities: 0.000 0.875 0.025 0.000 0.100
##
## Node number 6143: 6104 observations, complexity param=0.0007537688
## predicted class=Not.in.Labor.Force expected loss=0.06503932 P(node) =0.05785069
## class counts: 105 54 5707 8 230
## probabilities: 0.017 0.009 0.935 0.001 0.038
## left son=12286 (34 obs) right son=12287 (6070 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=61.524490, (0 missing)
## Industry.my.fctrPublic administration < 0.5 to the right, improve=21.341600, (0 missing)
## Industry.my.fctrMining < 0.5 to the left, improve=16.209240, (0 missing)
## Age < 17.5 to the left, improve= 9.827868, (0 missing)
##
## Node number 8190: 219 observations
## predicted class=Employed expected loss=0.06392694 P(node) =0.002075574
## class counts: 0 205 1 4 9
## probabilities: 0.000 0.936 0.005 0.018 0.041
##
## Node number 8191: 16912 observations, complexity param=0.001918684
## predicted class=Retired expected loss=0.09407521 P(node) =0.1602836
## class counts: 980 108 494 15321 9
## probabilities: 0.058 0.006 0.029 0.906 0.001
## left son=16382 (93 obs) right son=16383 (16819 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=152.67140, (0 missing)
## Age < 66.5 to the left, improve= 57.53166, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve= 38.20025, (0 missing)
##
## Node number 10238: 7944 observations, complexity param=0.008910715
## predicted class=Retired expected loss=0.6570997 P(node) =0.0752893
## class counts: 2647 853 1671 2724 49
## probabilities: 0.333 0.107 0.210 0.343 0.006
## left son=20476 (410 obs) right son=20477 (7534 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=428.8616, (0 missing)
## Industry.my.fctrInformation < 0.5 to the right, improve=329.0827, (0 missing)
## Age < 59.5 to the left, improve=280.2615, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=160.4271, (0 missing)
##
## Node number 10239: 11224 observations, complexity param=0.008910715
## predicted class=Not.in.Labor.Force expected loss=0.3833749 P(node) =0.1063755
## class counts: 1923 1677 6921 397 306
## probabilities: 0.171 0.149 0.617 0.035 0.027
## left son=20478 (870 obs) right son=20479 (10354 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=946.83090, (0 missing)
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=610.14300, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=386.43570, (0 missing)
## Age < 37.5 to the right, improve=187.39810, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 30.97126, (0 missing)
##
## Node number 12286: 34 observations
## predicted class=Employed expected loss=0.02941176 P(node) =0.0003222352
## class counts: 0 33 0 0 1
## probabilities: 0.000 0.971 0.000 0.000 0.029
##
## Node number 12287: 6070 observations, complexity param=0.0002512563
## predicted class=Not.in.Labor.Force expected loss=0.05980231 P(node) =0.05752846
## class counts: 105 21 5707 8 229
## probabilities: 0.017 0.003 0.940 0.001 0.038
## left son=24574 (18 obs) right son=24575 (6052 obs)
## Primary splits:
## Industry.my.fctrPublic administration < 0.5 to the right, improve=21.637840, (0 missing)
## Industry.my.fctrMining < 0.5 to the left, improve=16.401890, (0 missing)
## Age < 17.5 to the left, improve= 7.415681, (0 missing)
##
## Node number 16382: 93 observations
## predicted class=Employed expected loss=0.07526882 P(node) =0.000881408
## class counts: 0 86 0 2 5
## probabilities: 0.000 0.925 0.000 0.022 0.054
##
## Node number 16383: 16819 observations, complexity param=0.0001713111
## predicted class=Retired expected loss=0.08918485 P(node) =0.1594022
## class counts: 980 22 494 15319 4
## probabilities: 0.058 0.001 0.029 0.911 0.000
## left son=32766 (2744 obs) right son=32767 (14075 obs)
## Primary splits:
## Age < 66.5 to the left, improve=52.53577, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=38.64696, (0 missing)
##
## Node number 20476: 410 observations
## predicted class=Employed expected loss=0.04146341 P(node) =0.003885777
## class counts: 1 393 1 2 13
## probabilities: 0.002 0.959 0.002 0.005 0.032
##
## Node number 20477: 7534 observations, complexity param=0.008910715
## predicted class=Retired expected loss=0.6387045 P(node) =0.07140352
## class counts: 2646 460 1670 2722 36
## probabilities: 0.351 0.061 0.222 0.361 0.005
## left son=40954 (331 obs) right son=40955 (7203 obs)
## Primary splits:
## Industry.my.fctrInformation < 0.5 to the right, improve=367.4837, (0 missing)
## Age < 59.5 to the left, improve=288.8698, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=178.5305, (0 missing)
##
## Node number 20478: 870 observations
## predicted class=Employed expected loss=0.07701149 P(node) =0.008245429
## class counts: 0 803 5 1 61
## probabilities: 0.000 0.923 0.006 0.001 0.070
##
## Node number 20479: 10354 observations, complexity param=0.008910715
## predicted class=Not.in.Labor.Force expected loss=0.3320456 P(node) =0.09813009
## class counts: 1923 874 6916 396 245
## probabilities: 0.186 0.084 0.668 0.038 0.024
## left son=40958 (578 obs) right son=40959 (9776 obs)
## Primary splits:
## Industry.my.fctrAgriculture, forestry, fishing, and hunting < 0.5 to the right, improve=720.41770, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=453.50210, (0 missing)
## Age < 40.5 to the right, improve=193.70940, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 32.49911, (0 missing)
##
## Node number 24574: 18 observations
## predicted class=Employed expected loss=0.2222222 P(node) =0.0001705951
## class counts: 0 14 3 0 1
## probabilities: 0.000 0.778 0.167 0.000 0.056
##
## Node number 24575: 6052 observations, complexity param=0.0001598904
## predicted class=Not.in.Labor.Force expected loss=0.05750165 P(node) =0.05735786
## class counts: 105 7 5704 8 228
## probabilities: 0.017 0.001 0.942 0.001 0.038
## left son=49150 (6040 obs) right son=49151 (12 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the left, improve=16.486750, (0 missing)
## Age < 17.5 to the left, improve= 6.424928, (0 missing)
##
## Node number 32766: 2744 observations, complexity param=0.0001370489
## predicted class=Retired expected loss=0.1909621 P(node) =0.02600627
## class counts: 346 7 170 2220 1
## probabilities: 0.126 0.003 0.062 0.809 0.000
## left son=65532 (9 obs) right son=65533 (2735 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=9.990317, (0 missing)
## Age < 64.5 to the left, improve=8.016844, (0 missing)
##
## Node number 32767: 14075 observations, complexity param=0.0001713111
## predicted class=Retired expected loss=0.06934281 P(node) =0.1333959
## class counts: 634 15 324 13099 3
## probabilities: 0.045 0.001 0.023 0.931 0.000
## left son=65534 (17 obs) right son=65535 (14058 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=28.239480, (0 missing)
## Age < 70.5 to the left, improve= 7.452007, (0 missing)
##
## Node number 40954: 331 observations
## predicted class=Employed expected loss=0.0694864 P(node) =0.003137054
## class counts: 0 308 1 0 22
## probabilities: 0.000 0.931 0.003 0.000 0.066
##
## Node number 40955: 7203 observations, complexity param=0.008910715
## predicted class=Retired expected loss=0.6221019 P(node) =0.06826647
## class counts: 2646 152 1669 2722 14
## probabilities: 0.367 0.021 0.232 0.378 0.002
## left son=81910 (4514 obs) right son=81911 (2689 obs)
## Primary splits:
## Age < 59.5 to the left, improve=291.1724, (0 missing)
## Industry.my.fctrMining < 0.5 to the right, improve=195.1395, (0 missing)
##
## Node number 40958: 578 observations
## predicted class=Employed expected loss=0.07439446 P(node) =0.005477998
## class counts: 1 535 6 0 36
## probabilities: 0.002 0.926 0.010 0.000 0.062
##
## Node number 40959: 9776 observations, complexity param=0.007697579
## predicted class=Not.in.Labor.Force expected loss=0.2931669 P(node) =0.09265209
## class counts: 1922 339 6910 396 209
## probabilities: 0.197 0.035 0.707 0.041 0.021
## left son=81918 (354 obs) right son=81919 (9422 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=508.35280, (0 missing)
## Age < 40.5 to the right, improve=198.32880, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 33.88074, (0 missing)
##
## Node number 49150: 6040 observations
## predicted class=Not.in.Labor.Force expected loss=0.05562914 P(node) =0.05724413
## class counts: 105 0 5704 8 223
## probabilities: 0.017 0.000 0.944 0.001 0.037
##
## Node number 49151: 12 observations
## predicted class=Employed expected loss=0.4166667 P(node) =0.0001137301
## class counts: 0 7 0 0 5
## probabilities: 0.000 0.583 0.000 0.000 0.417
##
## Node number 65532: 9 observations
## predicted class=Employed expected loss=0.2222222 P(node) =8.529755e-05
## class counts: 0 7 1 1 0
## probabilities: 0.000 0.778 0.111 0.111 0.000
##
## Node number 65533: 2735 observations
## predicted class=Retired expected loss=0.1886654 P(node) =0.02592098
## class counts: 346 0 169 2219 1
## probabilities: 0.127 0.000 0.062 0.811 0.000
##
## Node number 65534: 17 observations
## predicted class=Employed expected loss=0.1176471 P(node) =0.0001611176
## class counts: 0 15 0 0 2
## probabilities: 0.000 0.882 0.000 0.000 0.118
##
## Node number 65535: 14058 observations
## predicted class=Retired expected loss=0.06821739 P(node) =0.1332348
## class counts: 634 0 324 13099 1
## probabilities: 0.045 0.000 0.023 0.932 0.000
##
## Node number 81910: 4514 observations, complexity param=0.002649612
## predicted class=Disabled expected loss=0.5750997 P(node) =0.04278146
## class counts: 1918 117 1342 1127 10
## probabilities: 0.425 0.026 0.297 0.250 0.002
## left son=163820 (123 obs) right son=163821 (4391 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=148.8923, (0 missing)
## Age < 54.5 to the left, improve= 43.2640, (0 missing)
##
## Node number 81911: 2689 observations, complexity param=0.0007994518
## predicted class=Retired expected loss=0.4068427 P(node) =0.02548501
## class counts: 728 35 327 1595 4
## probabilities: 0.271 0.013 0.122 0.593 0.001
## left son=163822 (35 obs) right son=163823 (2654 obs)
## Primary splits:
## Industry.my.fctrMining < 0.5 to the right, improve=50.14477, (0 missing)
## Age < 61.5 to the left, improve=21.72335, (0 missing)
##
## Node number 81918: 354 observations
## predicted class=Employed expected loss=0.04237288 P(node) =0.003355037
## class counts: 0 339 2 0 13
## probabilities: 0.000 0.958 0.006 0.000 0.037
##
## Node number 81919: 9422 observations, complexity param=0.0002055733
## predicted class=Not.in.Labor.Force expected loss=0.2668223 P(node) =0.08929705
## class counts: 1922 0 6908 396 196
## probabilities: 0.204 0.000 0.733 0.042 0.021
## left son=163838 (3482 obs) right son=163839 (5940 obs)
## Primary splits:
## Age < 37.5 to the right, improve=200.26230, (0 missing)
## Industry.my.fctrArmed forces < 0.5 to the left, improve= 34.89351, (0 missing)
##
## Node number 163820: 123 observations
## predicted class=Employed expected loss=0.04878049 P(node) =0.001165733
## class counts: 1 117 0 1 4
## probabilities: 0.008 0.951 0.000 0.008 0.033
##
## Node number 163821: 4391 observations
## predicted class=Disabled expected loss=0.5634252 P(node) =0.04161573
## class counts: 1917 0 1342 1126 6
## probabilities: 0.437 0.000 0.306 0.256 0.001
##
## Node number 163822: 35 observations
## predicted class=Employed expected loss=0 P(node) =0.0003317127
## class counts: 0 35 0 0 0
## probabilities: 0.000 1.000 0.000 0.000 0.000
##
## Node number 163823: 2654 observations
## predicted class=Retired expected loss=0.3990203 P(node) =0.0251533
## class counts: 728 0 327 1595 4
## probabilities: 0.274 0.000 0.123 0.601 0.002
##
## Node number 163838: 3482 observations
## predicted class=Not.in.Labor.Force expected loss=0.4029294 P(node) =0.03300067
## class counts: 1167 0 2079 211 25
## probabilities: 0.335 0.000 0.597 0.061 0.007
##
## Node number 163839: 5940 observations, complexity param=0.0002055733
## predicted class=Not.in.Labor.Force expected loss=0.187037 P(node) =0.05629638
## class counts: 755 0 4829 185 171
## probabilities: 0.127 0.000 0.813 0.031 0.029
## left son=327678 (5920 obs) right son=327679 (20 obs)
## Primary splits:
## Industry.my.fctrArmed forces < 0.5 to the left, improve=29.05522, (0 missing)
## Age < 27.5 to the right, improve=22.60554, (0 missing)
##
## Node number 327678: 5920 observations
## predicted class=Not.in.Labor.Force expected loss=0.1844595 P(node) =0.05610683
## class counts: 755 0 4828 185 152
## probabilities: 0.128 0.000 0.816 0.031 0.026
##
## Node number 327679: 20 observations
## predicted class=Unemployed expected loss=0.05 P(node) =0.0001895501
## class counts: 0 0 1 0 19
## probabilities: 0.000 0.000 0.050 0.000 0.950
##
## n= 105513
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 105513 43780 Employed (0.054 0.59 0.14 0.18 0.04)
## 2) Age< 63.5 83904 26650 Employed (0.056 0.68 0.18 0.038 0.048)
## 4) Age>=19.5 75489 20206 Employed (0.061 0.73 0.12 0.042 0.047)
## 8) Industry.my.fctrEducational and health services>=0.5 13669 684 Employed (0.00095 0.95 0.0048 0.0013 0.043) *
## 9) Industry.my.fctrEducational and health services< 0.5 61820 19522 Employed (0.075 0.68 0.14 0.051 0.047)
## 18) Industry.my.fctrTrade>=0.5 7743 578 Employed (0.0013 0.93 0.0065 0.00077 0.066) *
## 19) Industry.my.fctrTrade< 0.5 54077 18944 Employed (0.085 0.65 0.16 0.059 0.045)
## 38) Industry.my.fctrProfessional and business services>=0.5 6767 493 Employed (0.001 0.93 0.0052 0.00044 0.066) *
## 39) Industry.my.fctrProfessional and business services< 0.5 47310 18451 Employed (0.097 0.61 0.18 0.067 0.042)
## 78) Industry.my.fctrManufacturing>=0.5 6319 422 Employed (0.0013 0.93 0.0033 0.0016 0.061) *
## 79) Industry.my.fctrManufacturing< 0.5 40991 18029 Employed (0.11 0.56 0.21 0.077 0.039)
## 158) Industry.my.fctrLeisure and hospitality>=0.5 5109 463 Employed (0.00059 0.91 0.011 0.0012 0.078) *
## 159) Industry.my.fctrLeisure and hospitality< 0.5 35882 17566 Employed (0.13 0.51 0.24 0.088 0.033)
## 318) Industry.my.fctrFinancial>=0.5 3949 155 Employed (0 0.96 0.003 0.00051 0.036) *
## 319) Industry.my.fctrFinancial< 0.5 31933 17411 Employed (0.14 0.45 0.27 0.098 0.033)
## 638) Industry.my.fctrConstruction>=0.5 4073 330 Employed (0.00049 0.92 0.0047 0.0022 0.074) *
## 639) Industry.my.fctrConstruction< 0.5 27860 17081 Employed (0.16 0.39 0.31 0.11 0.027)
## 1278) Industry.my.fctrPublic administration>=0.5 2949 101 Employed (0.0014 0.97 0.0024 0.0027 0.028) *
## 1279) Industry.my.fctrPublic administration< 0.5 24911 16294 Not.in.Labor.Force (0.18 0.32 0.35 0.13 0.027)
## 2558) Industry.my.fctrTransportation and utilities>=0.5 2995 162 Employed (0.001 0.95 0.0027 0.0013 0.049) *
## 2559) Industry.my.fctrTransportation and utilities< 0.5 21916 13307 Not.in.Labor.Force (0.21 0.23 0.39 0.14 0.023)
## 5118) Industry.my.fctrOther services>=0.5 2748 180 Employed (0.00073 0.93 0.0062 0.00036 0.058) *
## 5119) Industry.my.fctrOther services< 0.5 19168 10576 Not.in.Labor.Force (0.24 0.13 0.45 0.16 0.019)
## 10238) Age>=49.5 7944 5220 Retired (0.33 0.11 0.21 0.34 0.0062)
## 20476) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 410 17 Employed (0.0024 0.96 0.0024 0.0049 0.032) *
## 20477) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 7534 4812 Retired (0.35 0.061 0.22 0.36 0.0048)
## 40954) Industry.my.fctrInformation>=0.5 331 23 Employed (0 0.93 0.003 0 0.066) *
## 40955) Industry.my.fctrInformation< 0.5 7203 4481 Retired (0.37 0.021 0.23 0.38 0.0019)
## 81910) Age< 59.5 4514 2596 Disabled (0.42 0.026 0.3 0.25 0.0022)
## 163820) Industry.my.fctrMining>=0.5 123 6 Employed (0.0081 0.95 0 0.0081 0.033) *
## 163821) Industry.my.fctrMining< 0.5 4391 2474 Disabled (0.44 0 0.31 0.26 0.0014) *
## 81911) Age>=59.5 2689 1094 Retired (0.27 0.013 0.12 0.59 0.0015)
## 163822) Industry.my.fctrMining>=0.5 35 0 Employed (0 1 0 0 0) *
## 163823) Industry.my.fctrMining< 0.5 2654 1059 Retired (0.27 0 0.12 0.6 0.0015) *
## 10239) Age< 49.5 11224 4303 Not.in.Labor.Force (0.17 0.15 0.62 0.035 0.027)
## 20478) Industry.my.fctrInformation>=0.5 870 67 Employed (0 0.92 0.0057 0.0011 0.07) *
## 20479) Industry.my.fctrInformation< 0.5 10354 3438 Not.in.Labor.Force (0.19 0.084 0.67 0.038 0.024)
## 40958) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 578 43 Employed (0.0017 0.93 0.01 0 0.062) *
## 40959) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 9776 2866 Not.in.Labor.Force (0.2 0.035 0.71 0.041 0.021)
## 81918) Industry.my.fctrMining>=0.5 354 15 Employed (0 0.96 0.0056 0 0.037) *
## 81919) Industry.my.fctrMining< 0.5 9422 2514 Not.in.Labor.Force (0.2 0 0.73 0.042 0.021)
## 163838) Age>=37.5 3482 1403 Not.in.Labor.Force (0.34 0 0.6 0.061 0.0072) *
## 163839) Age< 37.5 5940 1111 Not.in.Labor.Force (0.13 0 0.81 0.031 0.029)
## 327678) Industry.my.fctrArmed forces< 0.5 5920 1092 Not.in.Labor.Force (0.13 0 0.82 0.031 0.026) *
## 327679) Industry.my.fctrArmed forces>=0.5 20 1 Unemployed (0 0 0.05 0 0.95) *
## 5) Age< 19.5 8415 2554 Not.in.Labor.Force (0.012 0.23 0.7 0.00095 0.056)
## 10) Industry.my.fctrLeisure and hospitality>=0.5 930 136 Employed (0 0.85 0.063 0 0.083) *
## 11) Industry.my.fctrLeisure and hospitality< 0.5 7485 1683 Not.in.Labor.Force (0.014 0.16 0.78 0.0011 0.053)
## 22) Industry.my.fctrTrade>=0.5 543 74 Employed (0 0.86 0.037 0 0.099) *
## 23) Industry.my.fctrTrade< 0.5 6942 1160 Not.in.Labor.Force (0.015 0.1 0.83 0.0012 0.049)
## 46) Industry.my.fctrEducational and health services>=0.5 244 51 Employed (0 0.79 0.12 0 0.09) *
## 47) Industry.my.fctrEducational and health services< 0.5 6698 945 Not.in.Labor.Force (0.016 0.077 0.86 0.0012 0.047)
## 94) Industry.my.fctrOther services>=0.5 124 26 Employed (0 0.79 0.081 0 0.13) *
## 95) Industry.my.fctrOther services< 0.5 6574 831 Not.in.Labor.Force (0.016 0.063 0.87 0.0012 0.046)
## 190) Industry.my.fctrProfessional and business services>=0.5 128 38 Employed (0 0.7 0.13 0 0.17) *
## 191) Industry.my.fctrProfessional and business services< 0.5 6446 719 Not.in.Labor.Force (0.016 0.051 0.89 0.0012 0.043)
## 382) Industry.my.fctrConstruction>=0.5 89 16 Employed (0 0.82 0.067 0 0.11) *
## 383) Industry.my.fctrConstruction< 0.5 6357 636 Not.in.Labor.Force (0.017 0.04 0.9 0.0013 0.042)
## 766) Industry.my.fctrManufacturing>=0.5 89 24 Employed (0 0.73 0.045 0 0.22) *
## 767) Industry.my.fctrManufacturing< 0.5 6268 551 Not.in.Labor.Force (0.017 0.03 0.91 0.0013 0.04)
## 1534) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 76 20 Employed (0 0.74 0.11 0 0.16) *
## 1535) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 6192 483 Not.in.Labor.Force (0.017 0.021 0.92 0.0013 0.038)
## 3070) Industry.my.fctrFinancial>=0.5 48 4 Employed (0 0.92 0.021 0 0.063) *
## 3071) Industry.my.fctrFinancial< 0.5 6144 436 Not.in.Labor.Force (0.017 0.014 0.93 0.0013 0.038)
## 6142) Industry.my.fctrTransportation and utilities>=0.5 40 5 Employed (0 0.88 0.025 0 0.1) *
## 6143) Industry.my.fctrTransportation and utilities< 0.5 6104 397 Not.in.Labor.Force (0.017 0.0088 0.93 0.0013 0.038)
## 12286) Industry.my.fctrInformation>=0.5 34 1 Employed (0 0.97 0 0 0.029) *
## 12287) Industry.my.fctrInformation< 0.5 6070 363 Not.in.Labor.Force (0.017 0.0035 0.94 0.0013 0.038)
## 24574) Industry.my.fctrPublic administration>=0.5 18 4 Employed (0 0.78 0.17 0 0.056) *
## 24575) Industry.my.fctrPublic administration< 0.5 6052 348 Not.in.Labor.Force (0.017 0.0012 0.94 0.0013 0.038)
## 49150) Industry.my.fctrMining< 0.5 6040 336 Not.in.Labor.Force (0.017 0 0.94 0.0013 0.037) *
## 49151) Industry.my.fctrMining>=0.5 12 5 Employed (0 0.58 0 0 0.42) *
## 3) Age>=63.5 21609 6186 Retired (0.046 0.21 0.023 0.71 0.01)
## 6) Industry.my.fctrEducational and health services>=0.5 1104 78 Employed (0 0.93 0 0.028 0.043) *
## 7) Industry.my.fctrEducational and health services< 0.5 20505 5113 Retired (0.048 0.17 0.025 0.75 0.0083)
## 14) Industry.my.fctrTrade>=0.5 647 38 Employed (0.0046 0.94 0.0015 0.025 0.028) *
## 15) Industry.my.fctrTrade< 0.5 19858 4482 Retired (0.049 0.14 0.025 0.77 0.0077)
## 30) Industry.my.fctrProfessional and business services>=0.5 624 51 Employed (0.0016 0.92 0.0064 0.021 0.053) *
## 31) Industry.my.fctrProfessional and business services< 0.5 19234 3871 Retired (0.051 0.12 0.026 0.8 0.0062)
## 62) Industry.my.fctrManufacturing>=0.5 383 26 Employed (0 0.93 0.0052 0.01 0.052) *
## 63) Industry.my.fctrManufacturing< 0.5 18851 3492 Retired (0.052 0.1 0.026 0.81 0.0053)
## 126) Industry.my.fctrOther services>=0.5 352 21 Employed (0 0.94 0.0028 0.023 0.034) *
## 127) Industry.my.fctrOther services< 0.5 18499 3148 Retired (0.053 0.086 0.027 0.83 0.0047)
## 254) Industry.my.fctrFinancial>=0.5 350 27 Employed (0 0.92 0 0.023 0.054) *
## 255) Industry.my.fctrFinancial< 0.5 18149 2806 Retired (0.054 0.069 0.027 0.85 0.0037)
## 510) Industry.my.fctrLeisure and hospitality>=0.5 325 28 Employed (0.0031 0.91 0 0.022 0.062) *
## 511) Industry.my.fctrLeisure and hospitality< 0.5 17824 2488 Retired (0.055 0.054 0.028 0.86 0.0027)
## 1022) Industry.my.fctrAgriculture, forestry, fishing, and hunting>=0.5 243 8 Employed (0 0.97 0 0.012 0.021) *
## 1023) Industry.my.fctrAgriculture, forestry, fishing, and hunting< 0.5 17581 2248 Retired (0.056 0.041 0.028 0.87 0.0024)
## 2046) Industry.my.fctrConstruction>=0.5 225 18 Employed (0 0.92 0.0044 0.013 0.062) *
## 2047) Industry.my.fctrConstruction< 0.5 17356 2026 Retired (0.056 0.03 0.029 0.88 0.0017)
## 4094) Industry.my.fctrTransportation and utilities>=0.5 225 17 Employed (0 0.92 0.0044 0.022 0.049) *
## 4095) Industry.my.fctrTransportation and utilities< 0.5 17131 1806 Retired (0.057 0.018 0.029 0.89 0.0011)
## 8190) Industry.my.fctrPublic administration>=0.5 219 14 Employed (0 0.94 0.0046 0.018 0.041) *
## 8191) Industry.my.fctrPublic administration< 0.5 16912 1591 Retired (0.058 0.0064 0.029 0.91 0.00053)
## 16382) Industry.my.fctrInformation>=0.5 93 7 Employed (0 0.92 0 0.022 0.054) *
## 16383) Industry.my.fctrInformation< 0.5 16819 1500 Retired (0.058 0.0013 0.029 0.91 0.00024)
## 32766) Age< 66.5 2744 524 Retired (0.13 0.0026 0.062 0.81 0.00036)
## 65532) Industry.my.fctrMining>=0.5 9 2 Employed (0 0.78 0.11 0.11 0) *
## 65533) Industry.my.fctrMining< 0.5 2735 516 Retired (0.13 0 0.062 0.81 0.00037) *
## 32767) Age>=66.5 14075 976 Retired (0.045 0.0011 0.023 0.93 0.00021)
## 65534) Industry.my.fctrMining>=0.5 17 2 Employed (0 0.88 0 0 0.12) *
## 65535) Industry.my.fctrMining< 0.5 14058 959 Retired (0.045 0 0.023 0.93 7.1e-05) *
## [1] " calling mypredict_mdl for fit:"## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 1917 60 2027 1708
## Employed 0 61733 0 0
## Not.in.Labor.Force 1342 472 12611 820
## Retired 1126 176 404 16913
## Unemployed 6 3772 400 6
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 1
## Retired 0
## Unemployed 19
## Accuracy Kappa AccuracyLower AccuracyUpper AccuracyNull
## 0.8832371 NA 0.8812842 0.8851690 0.5850748
## AccuracyPValue McnemarPValue
## 0.0000000 0.0000000## Warning in ni[1:m] * nj[1:m]: NAs produced by integer overflow## Warning in mypredict_mdl(mdl, df = fit_df, rsp_var, rsp_var_out,
## model_id_method, : Expecting 1 metric: Accuracy; recd: Accuracy, Kappa;
## retaining Accuracy only
## model_id model_method feats max.nTuningRuns
## 1 Final.rpart rpart Industry.my.fctr, Age 1
## min.elapsedtime.everything min.elapsedtime.final max.Accuracy.fit
## 1 10.643 5.023 0.8827917
## max.AccuracyLower.fit max.AccuracyUpper.fit max.Kappa.fit
## 1 0.8812842 0.885169 0.79671
## max.AccuracySD.fit max.KappaSD.fit
## 1 0.001221035 0.002066722rm(ret_lst)
glb_chunks_df <- myadd_chunk(glb_chunks_df, "fit.data.training", major.inc=FALSE)## label step_major step_minor bgn end elapsed
## 14 fit.data.training 8 0 281.597 299.115 17.518
## 15 fit.data.training 8 1 299.115 NA NAglb_trnobs_df <- glb_get_predictions(df=glb_trnobs_df, mdl_id=glb_fin_mdl_id,
rsp_var_out=glb_rsp_var_out,
prob_threshold_def=ifelse(glb_is_classification && glb_is_binomial,
glb_models_df[glb_models_df$model_id == glb_sel_mdl_id, "opt.prob.threshold.OOB"], NULL))## Warning in `[<-.data.frame`(`*tmp*`, , paste0(rsp_var_out, ".prob"), value
## = structure(list(: provided 5 variables to replace 1 variablessav_featsimp_df <- glb_featsimp_df
#glb_feats_df <- sav_feats_df
# glb_feats_df <- mymerge_feats_importance(feats_df=glb_feats_df, sel_mdl=glb_fin_mdl,
# entity_df=glb_trnobs_df)
glb_featsimp_df <- myget_feats_importance(mdl=glb_fin_mdl, featsimp_df=glb_featsimp_df)
glb_featsimp_df[, paste0(glb_fin_mdl_id, ".importance")] <- glb_featsimp_df$importance
print(glb_featsimp_df)## Max.cor.Y.cv.0.cp.0.rpart.importance
## Age 100.000000
## Industry.my.fctrOther services 33.090343
## Industry.my.fctrEducational and health services 27.058509
## Industry.my.fctrTransportation and utilities 23.012814
## Industry.my.fctrManufacturing 28.370178
## Industry.my.fctrProfessional and business services 25.591436
## Industry.my.fctrTrade 24.329022
## Industry.my.fctrConstruction 22.685481
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 20.237759
## Industry.my.fctrPublic administration 29.566192
## Industry.my.fctrFinancial 21.645346
## Industry.my.fctrInformation 15.923595
## Industry.my.fctrLeisure and hospitality 22.062655
## Industry.my.fctrMining 11.722476
## Industry.my.fctrArmed forces 0.784697
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.000000
## `Industry.my.fctrArmed forces` 0.000000
## `Industry.my.fctrEducational and health services` 0.000000
## `Industry.my.fctrLeisure and hospitality` 0.000000
## `Industry.my.fctrOther services` 0.000000
## `Industry.my.fctrProfessional and business services` 0.000000
## `Industry.my.fctrPublic administration` 0.000000
## `Industry.my.fctrTransportation and utilities` 0.000000
## importance
## Age 100.0000000
## Industry.my.fctrOther services 27.4614104
## Industry.my.fctrEducational and health services 25.2415079
## Industry.my.fctrTransportation and utilities 23.7839646
## Industry.my.fctrManufacturing 23.1388463
## Industry.my.fctrProfessional and business services 21.2185994
## Industry.my.fctrTrade 20.1374624
## Industry.my.fctrConstruction 20.0204644
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 19.9699985
## Industry.my.fctrPublic administration 19.9449139
## Industry.my.fctrFinancial 19.7491177
## Industry.my.fctrInformation 19.2223509
## Industry.my.fctrLeisure and hospitality 16.1591357
## Industry.my.fctrMining 10.4806617
## Industry.my.fctrArmed forces 0.6223753
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.0000000
## `Industry.my.fctrArmed forces` 0.0000000
## `Industry.my.fctrEducational and health services` 0.0000000
## `Industry.my.fctrLeisure and hospitality` 0.0000000
## `Industry.my.fctrOther services` 0.0000000
## `Industry.my.fctrProfessional and business services` 0.0000000
## `Industry.my.fctrPublic administration` 0.0000000
## `Industry.my.fctrTransportation and utilities` 0.0000000
## Final.rpart.importance
## Age 100.0000000
## Industry.my.fctrOther services 27.4614104
## Industry.my.fctrEducational and health services 25.2415079
## Industry.my.fctrTransportation and utilities 23.7839646
## Industry.my.fctrManufacturing 23.1388463
## Industry.my.fctrProfessional and business services 21.2185994
## Industry.my.fctrTrade 20.1374624
## Industry.my.fctrConstruction 20.0204644
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 19.9699985
## Industry.my.fctrPublic administration 19.9449139
## Industry.my.fctrFinancial 19.7491177
## Industry.my.fctrInformation 19.2223509
## Industry.my.fctrLeisure and hospitality 16.1591357
## Industry.my.fctrMining 10.4806617
## Industry.my.fctrArmed forces 0.6223753
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.0000000
## `Industry.my.fctrArmed forces` 0.0000000
## `Industry.my.fctrEducational and health services` 0.0000000
## `Industry.my.fctrLeisure and hospitality` 0.0000000
## `Industry.my.fctrOther services` 0.0000000
## `Industry.my.fctrProfessional and business services` 0.0000000
## `Industry.my.fctrPublic administration` 0.0000000
## `Industry.my.fctrTransportation and utilities` 0.0000000if (glb_is_classification && glb_is_binomial)
glb_analytics_diag_plots(obs_df=glb_trnobs_df, mdl_id=glb_fin_mdl_id,
prob_threshold=glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"opt.prob.threshold.OOB"]) else
glb_analytics_diag_plots(obs_df=glb_trnobs_df, mdl_id=glb_fin_mdl_id) 
## [1] "Min/Max Boundaries: "
## .rownames EmploymentStatus.fctr
## 10 93435 Not.in.Labor.Force
## 107 35055 Retired
## 910 80194 Disabled
## 4110 111997 Disabled
## 5360 55065 Disabled
## 5476 55388 Disabled
## 5740 57377 Not.in.Labor.Force
## 7089 664 Not.in.Labor.Force
## 3453 34149 Disabled
## 3898 33583 Disabled
## 12742 35001 Unemployed
## 13154 41299 Disabled
## 5003 57040 Not.in.Labor.Force
## 8217 20026 Not.in.Labor.Force
## EmploymentStatus.fctr.predict.Final.rpart
## 10 Not.in.Labor.Force
## 107 Retired
## 910 Retired
## 4110 Retired
## 5360 Retired
## 5476 Retired
## 5740 Retired
## 7089 Retired
## 3453 Not.in.Labor.Force
## 3898 Not.in.Labor.Force
## 12742 Not.in.Labor.Force
## 13154 Not.in.Labor.Force
## 5003 Employed
## 8217 Employed
## EmploymentStatus.fctr.predict.Final.rpart.prob
## 10 0.01738411
## 107 0.04509888
## 910 0.04509888
## 4110 0.04509888
## 5360 0.04509888
## 5476 0.04509888
## 5740 0.04509888
## 7089 0.04509888
## 3453 0.01738411
## 3898 0.01738411
## 12742 0.01738411
## 13154 0.01738411
## 5003 0.00000000
## 8217 0.00000000
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## 10 TRUE
## 107 TRUE
## 910 FALSE
## 4110 FALSE
## 5360 FALSE
## 5476 FALSE
## 5740 FALSE
## 7089 FALSE
## 3453 FALSE
## 3898 FALSE
## 12742 FALSE
## 13154 FALSE
## 5003 FALSE
## 8217 FALSE
## EmploymentStatus.fctr.predict.Final.rpart.error .label
## 10 0.0000000 93435
## 107 0.0000000 35055
## 910 0.9549011 80194
## 4110 0.9549011 111997
## 5360 0.9549011 55065
## 5476 0.9549011 55388
## 5740 0.9549011 57377
## 7089 0.9549011 664
## 3453 0.9826159 34149
## 3898 0.9826159 33583
## 12742 0.9826159 35001
## 13154 0.9826159 41299
## 5003 1.0000000 57040
## 8217 1.0000000 20026
## [1] "Inaccurate: "
## .rownames EmploymentStatus.fctr
## 89 91014 Not.in.Labor.Force
## 93 92066 Not.in.Labor.Force
## 119 25235 Retired
## 131 34884 Retired
## 148 92331 Not.in.Labor.Force
## 161 24269 Retired
## EmploymentStatus.fctr.predict.Final.rpart
## 89 Disabled
## 93 Disabled
## 119 Disabled
## 131 Disabled
## 148 Disabled
## 161 Disabled
## EmploymentStatus.fctr.predict.Final.rpart.prob
## 89 0.4365748
## 93 0.4365748
## 119 0.4365748
## 131 0.4365748
## 148 0.4365748
## 161 0.4365748
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## 89 FALSE
## 93 FALSE
## 119 FALSE
## 131 FALSE
## 148 FALSE
## 161 FALSE
## EmploymentStatus.fctr.predict.Final.rpart.error
## 89 0.5634252
## 93 0.5634252
## 119 0.5634252
## 131 0.5634252
## 148 0.5634252
## 161 0.5634252
## .rownames EmploymentStatus.fctr
## 130472 81794 Retired
## 105398 95771 Disabled
## 82516 27086 Disabled
## 83977 2630 Unemployed
## 60583 25168 Disabled
## 47827 107492 Unemployed
## EmploymentStatus.fctr.predict.Final.rpart
## 130472 Disabled
## 105398 Not.in.Labor.Force
## 82516 Retired
## 83977 Not.in.Labor.Force
## 60583 Retired
## 47827 Employed
## EmploymentStatus.fctr.predict.Final.rpart.prob
## 130472 0.436574812
## 105398 0.335152211
## 82516 0.274302939
## 83977 0.127533784
## 60583 0.045098876
## 47827 0.001291489
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## 130472 FALSE
## 105398 FALSE
## 82516 FALSE
## 83977 FALSE
## 60583 FALSE
## 47827 FALSE
## EmploymentStatus.fctr.predict.Final.rpart.error
## 130472 0.5634252
## 105398 0.6648478
## 82516 0.7256971
## 83977 0.8724662
## 60583 0.9549011
## 47827 0.9987085
## .rownames EmploymentStatus.fctr
## 130654 79473 Unemployed
## 130684 95119 Unemployed
## 130736 20447 Unemployed
## 130840 97074 Unemployed
## 130934 33835 Unemployed
## 131021 17653 Unemployed
## EmploymentStatus.fctr.predict.Final.rpart
## 130654 Employed
## 130684 Employed
## 130736 Employed
## 130840 Employed
## 130934 Employed
## 131021 Employed
## EmploymentStatus.fctr.predict.Final.rpart.prob
## 130654 0
## 130684 0
## 130736 0
## 130840 0
## 130934 0
## 131021 0
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## 130654 FALSE
## 130684 FALSE
## 130736 FALSE
## 130840 FALSE
## 130934 FALSE
## 131021 FALSE
## EmploymentStatus.fctr.predict.Final.rpart.error
## 130654 1
## 130684 1
## 130736 1
## 130840 1
## 130934 1
## 131021 1
dsp_feats_vctr <- c(NULL)
for(var in grep(".importance", names(glb_feats_df), fixed=TRUE, value=TRUE))
dsp_feats_vctr <- union(dsp_feats_vctr,
glb_feats_df[!is.na(glb_feats_df[, var]), "id"])
print(glb_trnobs_df[glb_trnobs_df$UniqueID %in% FN_OOB_ids,
grep(glb_rsp_var, names(glb_trnobs_df), value=TRUE)])## [1] EmploymentStatus.fctr
## [2] EmploymentStatus.fctr.predict.Final.rpart
## [3] EmploymentStatus.fctr.predict.Final.rpart.prob
## <0 rows> (or 0-length row.names)print(setdiff(names(glb_trnobs_df), names(glb_allobs_df)))## [1] "EmploymentStatus.fctr.predict.Final.rpart"
## [2] "EmploymentStatus.fctr.predict.Final.rpart.prob"for (col in setdiff(names(glb_trnobs_df), names(glb_allobs_df)))
# Merge or cbind ?
glb_allobs_df[glb_allobs_df$.src == "Train", col] <- glb_trnobs_df[, col]
print(setdiff(names(glb_fitobs_df), names(glb_allobs_df)))## character(0)print(setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))## character(0)for (col in setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))
# Merge or cbind ?
glb_allobs_df[glb_allobs_df$.lcn == "OOB", col] <- glb_OOBobs_df[, col]
print(setdiff(names(glb_newobs_df), names(glb_allobs_df)))## character(0)if (glb_save_envir)
save(glb_feats_df, glb_allobs_df,
#glb_trnobs_df, glb_fitobs_df, glb_OOBobs_df, glb_newobs_df,
glb_models_df, dsp_models_df, glb_models_lst, glb_model_type,
glb_sel_mdl, glb_sel_mdl_id,
glb_fin_mdl, glb_fin_mdl_id,
file=paste0(glb_out_pfx, "dsk.RData"))
replay.petrisim(pn=glb_analytics_pn,
replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
"data.training.all.prediction","model.final")), flip_coord=TRUE)## time trans "bgn " "fit.data.training.all " "predict.data.new " "end "
## 0.0000 multiple enabled transitions: data.training.all data.new model.selected firing: data.training.all
## 1.0000 1 2 1 0 0
## 1.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction firing: data.new
## 2.0000 2 1 1 1 0
## 2.0000 multiple enabled transitions: data.training.all data.new model.selected model.final data.training.all.prediction data.new.prediction firing: model.selected
## 3.0000 3 0 2 1 0
## 3.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: data.training.all.prediction
## 4.0000 5 0 1 1 1
## 4.0000 multiple enabled transitions: model.final data.training.all.prediction data.new.prediction firing: model.final
## 5.0000 4 0 0 2 1
glb_chunks_df <- myadd_chunk(glb_chunks_df, "predict.data.new", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 15 fit.data.training 8 1 299.115 345.244 46.129
## 16 predict.data.new 9 0 345.244 NA NAStep 9.0: predict data new
# Compute final model predictions
glb_newobs_df <- glb_get_predictions(glb_newobs_df, mdl_id=glb_fin_mdl_id,
rsp_var_out=glb_rsp_var_out,
prob_threshold_def=ifelse(glb_is_classification && glb_is_binomial,
glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"opt.prob.threshold.OOB"], NULL))## Warning in `[<-.data.frame`(`*tmp*`, , paste0(rsp_var_out, ".prob"), value
## = structure(list(: provided 5 variables to replace 1 variablesif (glb_is_classification && glb_is_binomial)
glb_analytics_diag_plots(obs_df=glb_newobs_df, mdl_id=glb_fin_mdl_id,
prob_threshold=glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"opt.prob.threshold.OOB"]) else
glb_analytics_diag_plots(obs_df=glb_newobs_df, mdl_id=glb_fin_mdl_id) ## Warning in loop_apply(n, do.ply): no non-missing arguments to min;
## returning Inf## Warning in loop_apply(n, do.ply): no non-missing arguments to max;
## returning -Inf## Warning in loop_apply(n, do.ply): Removed 25789 rows containing missing
## values (geom_point).## Warning in loop_apply(n, do.ply): Removed 25789 rows containing missing
## values (geom_point).
## Warning in loop_apply(n, do.ply): no non-missing arguments to min;
## returning Inf## Warning in loop_apply(n, do.ply): no non-missing arguments to max;
## returning -Inf## Warning in loop_apply(n, do.ply): Removed 25789 rows containing missing
## values (geom_point).## Warning in loop_apply(n, do.ply): Removed 25789 rows containing missing
## values (geom_point).
## [1] "Min/Max Boundaries: "
## .rownames EmploymentStatus.fctr
## 26 92346 <NA>
## 12016 117703 <NA>
## EmploymentStatus.fctr.predict.Final.rpart
## 26 Not.in.Labor.Force
## 12016 Retired
## EmploymentStatus.fctr.predict.Final.rpart.prob
## 26 0.01738411
## 12016 0.27430294
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## 26 NA
## 12016 NA
## EmploymentStatus.fctr.predict.Final.rpart.error .label
## 26 0 92346
## 12016 0 117703
## [1] "Inaccurate: "
## .rownames EmploymentStatus.fctr
## NA <NA> <NA>
## NA.1 <NA> <NA>
## NA.2 <NA> <NA>
## NA.3 <NA> <NA>
## NA.4 <NA> <NA>
## NA.5 <NA> <NA>
## EmploymentStatus.fctr.predict.Final.rpart
## NA <NA>
## NA.1 <NA>
## NA.2 <NA>
## NA.3 <NA>
## NA.4 <NA>
## NA.5 <NA>
## EmploymentStatus.fctr.predict.Final.rpart.prob
## NA NA
## NA.1 NA
## NA.2 NA
## NA.3 NA
## NA.4 NA
## NA.5 NA
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## NA NA
## NA.1 NA
## NA.2 NA
## NA.3 NA
## NA.4 NA
## NA.5 NA
## EmploymentStatus.fctr.predict.Final.rpart.error
## NA NA
## NA.1 NA
## NA.2 NA
## NA.3 NA
## NA.4 NA
## NA.5 NA
## .rownames EmploymentStatus.fctr
## NA.651 <NA> <NA>
## NA.7634 <NA> <NA>
## NA.11771 <NA> <NA>
## NA.14528 <NA> <NA>
## NA.21351 <NA> <NA>
## NA.23439 <NA> <NA>
## EmploymentStatus.fctr.predict.Final.rpart
## NA.651 <NA>
## NA.7634 <NA>
## NA.11771 <NA>
## NA.14528 <NA>
## NA.21351 <NA>
## NA.23439 <NA>
## EmploymentStatus.fctr.predict.Final.rpart.prob
## NA.651 NA
## NA.7634 NA
## NA.11771 NA
## NA.14528 NA
## NA.21351 NA
## NA.23439 NA
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## NA.651 NA
## NA.7634 NA
## NA.11771 NA
## NA.14528 NA
## NA.21351 NA
## NA.23439 NA
## EmploymentStatus.fctr.predict.Final.rpart.error
## NA.651 NA
## NA.7634 NA
## NA.11771 NA
## NA.14528 NA
## NA.21351 NA
## NA.23439 NA
## .rownames EmploymentStatus.fctr
## NA.25783 <NA> <NA>
## NA.25784 <NA> <NA>
## NA.25785 <NA> <NA>
## NA.25786 <NA> <NA>
## NA.25787 <NA> <NA>
## NA.25788 <NA> <NA>
## EmploymentStatus.fctr.predict.Final.rpart
## NA.25783 <NA>
## NA.25784 <NA>
## NA.25785 <NA>
## NA.25786 <NA>
## NA.25787 <NA>
## NA.25788 <NA>
## EmploymentStatus.fctr.predict.Final.rpart.prob
## NA.25783 NA
## NA.25784 NA
## NA.25785 NA
## NA.25786 NA
## NA.25787 NA
## NA.25788 NA
## EmploymentStatus.fctr.predict.Final.rpart.accurate
## NA.25783 NA
## NA.25784 NA
## NA.25785 NA
## NA.25786 NA
## NA.25787 NA
## NA.25788 NA
## EmploymentStatus.fctr.predict.Final.rpart.error
## NA.25783 NA
## NA.25784 NA
## NA.25785 NA
## NA.25786 NA
## NA.25787 NA
## NA.25788 NA## Warning in loop_apply(n, do.ply): Removed 25789 rows containing missing
## values (geom_point).
if (glb_is_classification && glb_is_binomial) {
submit_df <- glb_newobs_df[, c(glb_id_var,
paste0(glb_rsp_var_out, glb_fin_mdl_id, ".prob"))]
names(submit_df)[2] <- "Probability1"
} else submit_df <- glb_newobs_df[, c(glb_id_var,
paste0(glb_rsp_var_out, glb_fin_mdl_id))]
write.csv(submit_df,
paste0(gsub(".", "_", paste0(glb_out_pfx, glb_fin_mdl_id), fixed=TRUE),
"_submit.csv"), row.names=FALSE)
# print(orderBy(~ -max.auc.OOB, glb_models_df[, c("model_id",
# "max.auc.OOB", "max.Accuracy.OOB")]))
if (glb_is_classification && glb_is_binomial)
print(glb_models_df[glb_models_df$model_id == glb_sel_mdl_id,
"opt.prob.threshold.OOB"])
print(sprintf("glb_sel_mdl_id: %s", glb_sel_mdl_id))## [1] "glb_sel_mdl_id: Max.cor.Y.cv.0.cp.0.rpart"print(sprintf("glb_fin_mdl_id: %s", glb_fin_mdl_id))## [1] "glb_fin_mdl_id: Final.rpart"print(dim(glb_fitobs_df))## [1] 77145 35print(dsp_models_df)## model_id max.Accuracy.OOB max.Kappa.OOB
## 4 Max.cor.Y.cv.0.cp.0.rpart 0.8828257 0.7967286427
## 5 Max.cor.Y.rpart 0.7261351 0.4752624489
## 6 Low.cor.X.rpart 0.7261351 0.4752624489
## 7 All.X.no.rnorm.rpart 0.7261351 0.4752624489
## 1 MFO.myMFO_classfr 0.5850606 0.0000000000
## 3 Max.cor.Y.cv.0.rpart 0.5850606 0.0000000000
## 2 Random.myrandom_classfr 0.3982657 -0.0008038519if (glb_is_classification) {
print(sprintf("%s OOB confusion matrix & accuracy: ", glb_sel_mdl_id))
print(t(confusionMatrix(glb_OOBobs_df[, paste0(glb_rsp_var_out, glb_sel_mdl_id)],
glb_OOBobs_df[, glb_rsp_var])$table))
if (!is.null(glb_category_vars)) {
tmp_OOBobs_df <- glb_OOBobs_df[, c(glb_category_vars, predct_accurate_var_name)]
names(tmp_OOBobs_df)[length(names(tmp_OOBobs_df))] <- "accurate.OOB"
aOOB_ctgry_df <- mycreate_xtab_df(tmp_OOBobs_df, names(tmp_OOBobs_df))
aOOB_ctgry_df[is.na(aOOB_ctgry_df)] <- 0
aOOB_ctgry_df <- mutate(aOOB_ctgry_df,
.n.OOB = accurate.OOB.FALSE + accurate.OOB.TRUE,
max.accuracy.OOB = accurate.OOB.TRUE / .n.OOB)
#intersect(names(glb_ctgry_df), names(aOOB_ctgry_df))
glb_ctgry_df <- merge(glb_ctgry_df, aOOB_ctgry_df, all=TRUE)
print(orderBy(~-accurate.OOB.FALSE, glb_ctgry_df))
}
} ## [1] "Max.cor.Y.cv.0.cp.0.rpart OOB confusion matrix & accuracy: "
## Prediction
## Reference Disabled Employed Not.in.Labor.Force Retired
## Disabled 468 24 534 510
## Employed 0 16594 1 2
## Not.in.Labor.Force 348 133 3369 248
## Retired 244 41 111 4610
## Unemployed 3 1006 117 1
## Prediction
## Reference Unemployed
## Disabled 0
## Employed 0
## Not.in.Labor.Force 1
## Retired 0
## Unemployed 3dsp_myCategory_conf_mtrx <- function(myCategory) {
print(sprintf("%s OOB::myCategory=%s confusion matrix & accuracy: ",
glb_sel_mdl_id, myCategory))
print(t(confusionMatrix(
glb_OOBobs_df[glb_OOBobs_df$myCategory == myCategory,
paste0(glb_rsp_var_out, glb_sel_mdl_id)],
glb_OOBobs_df[glb_OOBobs_df$myCategory == myCategory, glb_rsp_var])$table))
print(sum(glb_OOBobs_df[glb_OOBobs_df$myCategory == myCategory,
predct_accurate_var_name]) /
nrow(glb_OOBobs_df[glb_OOBobs_df$myCategory == myCategory, ]))
err_ids <- glb_OOBobs_df[(glb_OOBobs_df$myCategory == myCategory) &
(!glb_OOBobs_df[, predct_accurate_var_name]), glb_id_var]
OOB_FNerr_df <- glb_OOBobs_df[(glb_OOBobs_df$UniqueID %in% err_ids) &
(glb_OOBobs_df$Popular == 1),
c(
".clusterid",
"Popular", "Headline", "Snippet", "Abstract")]
print(sprintf("%s OOB::myCategory=%s FN errors: %d", glb_sel_mdl_id, myCategory,
nrow(OOB_FNerr_df)))
print(OOB_FNerr_df)
OOB_FPerr_df <- glb_OOBobs_df[(glb_OOBobs_df$UniqueID %in% err_ids) &
(glb_OOBobs_df$Popular == 0),
c(
".clusterid",
"Popular", "Headline", "Snippet", "Abstract")]
print(sprintf("%s OOB::myCategory=%s FP errors: %d", glb_sel_mdl_id, myCategory,
nrow(OOB_FPerr_df)))
print(OOB_FPerr_df)
}
#dsp_myCategory_conf_mtrx(myCategory="OpEd#Opinion#")
#dsp_myCategory_conf_mtrx(myCategory="Business#Business Day#Dealbook")
#dsp_myCategory_conf_mtrx(myCategory="##")
if (glb_is_classification) {
print("FN_OOB_ids:")
print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
grep(glb_rsp_var, names(glb_OOBobs_df), value=TRUE)])
print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
glb_txt_vars])
print(dsp_vctr <- colSums(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
setdiff(grep("[HSA].", names(glb_OOBobs_df), value=TRUE),
union(myfind_chr_cols_df(glb_OOBobs_df),
grep(".fctr", names(glb_OOBobs_df), fixed=TRUE, value=TRUE)))]))
}## [1] "FN_OOB_ids:"
## [1] EmploymentStatus.fctr
## [2] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart
## [3] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.prob
## [4] EmploymentStatus.fctr.predict.Max.cor.Y.cv.0.cp.0.rpart.accurate
## <0 rows> (or 0-length row.names)
## data frame with 0 columns and 0 rows
## MetroAreaCode PeopleInHousehold Age Hispanic
## 0 0 0 0dsp_hdlpfx_results <- function(hdlpfx) {
print(hdlpfx)
print(glb_OOBobs_df[glb_OOBobs_df$Headline.pfx %in% c(hdlpfx),
grep(glb_rsp_var, names(glb_OOBobs_df), value=TRUE)])
print(glb_newobs_df[glb_newobs_df$Headline.pfx %in% c(hdlpfx),
grep(glb_rsp_var, names(glb_newobs_df), value=TRUE)])
print(dsp_vctr <- colSums(glb_newobs_df[glb_newobs_df$Headline.pfx %in% c(hdlpfx),
setdiff(grep("[HSA]\\.", names(glb_newobs_df), value=TRUE),
union(myfind_chr_cols_df(glb_newobs_df),
grep(".fctr", names(glb_newobs_df), fixed=TRUE, value=TRUE)))]))
print(dsp_vctr <- dsp_vctr[dsp_vctr != 0])
print(glb_newobs_df[glb_newobs_df$Headline.pfx %in% c(hdlpfx),
union(names(dsp_vctr), myfind_chr_cols_df(glb_newobs_df))])
}
#dsp_hdlpfx_results(hdlpfx="Ask Well::")
# print("myMisc::|OpEd|blank|blank|1:")
# print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% c(6446),
# grep(glb_rsp_var, names(glb_OOBobs_df), value=TRUE)])
# print(glb_OOBobs_df[glb_OOBobs_df$UniqueID %in% FN_OOB_ids,
# c("WordCount", "WordCount.log", "myMultimedia",
# "NewsDesk", "SectionName", "SubsectionName")])
# print(mycreate_sqlxtab_df(glb_allobs_df[sel_obs(Headline.contains="[Vv]ideo"), ],
# c(glb_rsp_var, "myMultimedia")))
# dsp_chisq.test(Headline.contains="[Vi]deo")
# print(glb_allobs_df[sel_obs(Headline.contains="[Vv]ideo"),
# c(glb_rsp_var, "Popular", "myMultimedia", "Headline")])
# print(glb_allobs_df[sel_obs(Headline.contains="[Ee]bola", Popular=1),
# c(glb_rsp_var, "Popular", "myMultimedia", "Headline",
# "NewsDesk", "SectionName", "SubsectionName")])
# print(subset(glb_feats_df, !is.na(importance))[,
# c("is.ConditionalX.y",
# grep("importance", names(glb_feats_df), fixed=TRUE, value=TRUE))])
# print(subset(glb_feats_df, is.ConditionalX.y & is.na(importance))[,
# c("is.ConditionalX.y",
# grep("importance", names(glb_feats_df), fixed=TRUE, value=TRUE))])
# print(subset(glb_feats_df, !is.na(importance))[,
# c("zeroVar", "nzv", "myNearZV",
# grep("importance", names(glb_feats_df), fixed=TRUE, value=TRUE))])
# print(subset(glb_feats_df, is.na(importance))[,
# c("zeroVar", "nzv", "myNearZV",
# grep("importance", names(glb_feats_df), fixed=TRUE, value=TRUE))])
print(orderBy(as.formula(paste0("~ -", glb_sel_mdl_id, ".importance")), glb_featsimp_df))## Max.cor.Y.cv.0.cp.0.rpart.importance
## Age 100.000000
## Industry.my.fctrOther services 33.090343
## Industry.my.fctrPublic administration 29.566192
## Industry.my.fctrManufacturing 28.370178
## Industry.my.fctrEducational and health services 27.058509
## Industry.my.fctrProfessional and business services 25.591436
## Industry.my.fctrTrade 24.329022
## Industry.my.fctrTransportation and utilities 23.012814
## Industry.my.fctrConstruction 22.685481
## Industry.my.fctrLeisure and hospitality 22.062655
## Industry.my.fctrFinancial 21.645346
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 20.237759
## Industry.my.fctrInformation 15.923595
## Industry.my.fctrMining 11.722476
## Industry.my.fctrArmed forces 0.784697
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.000000
## `Industry.my.fctrArmed forces` 0.000000
## `Industry.my.fctrEducational and health services` 0.000000
## `Industry.my.fctrLeisure and hospitality` 0.000000
## `Industry.my.fctrOther services` 0.000000
## `Industry.my.fctrProfessional and business services` 0.000000
## `Industry.my.fctrPublic administration` 0.000000
## `Industry.my.fctrTransportation and utilities` 0.000000
## importance
## Age 100.0000000
## Industry.my.fctrOther services 27.4614104
## Industry.my.fctrPublic administration 19.9449139
## Industry.my.fctrManufacturing 23.1388463
## Industry.my.fctrEducational and health services 25.2415079
## Industry.my.fctrProfessional and business services 21.2185994
## Industry.my.fctrTrade 20.1374624
## Industry.my.fctrTransportation and utilities 23.7839646
## Industry.my.fctrConstruction 20.0204644
## Industry.my.fctrLeisure and hospitality 16.1591357
## Industry.my.fctrFinancial 19.7491177
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 19.9699985
## Industry.my.fctrInformation 19.2223509
## Industry.my.fctrMining 10.4806617
## Industry.my.fctrArmed forces 0.6223753
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.0000000
## `Industry.my.fctrArmed forces` 0.0000000
## `Industry.my.fctrEducational and health services` 0.0000000
## `Industry.my.fctrLeisure and hospitality` 0.0000000
## `Industry.my.fctrOther services` 0.0000000
## `Industry.my.fctrProfessional and business services` 0.0000000
## `Industry.my.fctrPublic administration` 0.0000000
## `Industry.my.fctrTransportation and utilities` 0.0000000
## Final.rpart.importance
## Age 100.0000000
## Industry.my.fctrOther services 27.4614104
## Industry.my.fctrPublic administration 19.9449139
## Industry.my.fctrManufacturing 23.1388463
## Industry.my.fctrEducational and health services 25.2415079
## Industry.my.fctrProfessional and business services 21.2185994
## Industry.my.fctrTrade 20.1374624
## Industry.my.fctrTransportation and utilities 23.7839646
## Industry.my.fctrConstruction 20.0204644
## Industry.my.fctrLeisure and hospitality 16.1591357
## Industry.my.fctrFinancial 19.7491177
## Industry.my.fctrAgriculture, forestry, fishing, and hunting 19.9699985
## Industry.my.fctrInformation 19.2223509
## Industry.my.fctrMining 10.4806617
## Industry.my.fctrArmed forces 0.6223753
## `Industry.my.fctrAgriculture, forestry, fishing, and hunting` 0.0000000
## `Industry.my.fctrArmed forces` 0.0000000
## `Industry.my.fctrEducational and health services` 0.0000000
## `Industry.my.fctrLeisure and hospitality` 0.0000000
## `Industry.my.fctrOther services` 0.0000000
## `Industry.my.fctrProfessional and business services` 0.0000000
## `Industry.my.fctrPublic administration` 0.0000000
## `Industry.my.fctrTransportation and utilities` 0.0000000print(setdiff(names(glb_trnobs_df), names(glb_allobs_df)))## character(0)for (col in setdiff(names(glb_trnobs_df), names(glb_allobs_df)))
# Merge or cbind ?
glb_allobs_df[glb_allobs_df$.src == "Train", col] <- glb_trnobs_df[, col]
print(setdiff(names(glb_fitobs_df), names(glb_allobs_df)))## character(0)print(setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))## character(0)for (col in setdiff(names(glb_OOBobs_df), names(glb_allobs_df)))
# Merge or cbind ?
glb_allobs_df[glb_allobs_df$.lcn == "OOB", col] <- glb_OOBobs_df[, col]
print(setdiff(names(glb_newobs_df), names(glb_allobs_df)))## character(0)if (glb_save_envir)
save(glb_feats_df, glb_allobs_df,
#glb_trnobs_df, glb_fitobs_df, glb_OOBobs_df, glb_newobs_df,
glb_models_df, dsp_models_df, glb_models_lst, glb_model_type,
glb_sel_mdl, glb_sel_mdl_id,
glb_fin_mdl, glb_fin_mdl_id,
file=paste0(glb_out_pfx, "prdnew_dsk.RData"))
rm(submit_df, tmp_OOBobs_df)## Warning in rm(submit_df, tmp_OOBobs_df): object 'tmp_OOBobs_df' not found# tmp_replay_lst <- replay.petrisim(pn=glb_analytics_pn,
# replay.trans=(glb_analytics_avl_objs <- c(glb_analytics_avl_objs,
# "data.new.prediction")), flip_coord=TRUE)
# print(ggplot.petrinet(tmp_replay_lst[["pn"]]) + coord_flip())
glb_chunks_df <- myadd_chunk(glb_chunks_df, "display.session.info", major.inc=TRUE)## label step_major step_minor bgn end elapsed
## 16 predict.data.new 9 0 345.244 382.729 37.485
## 17 display.session.info 10 0 382.729 NA NANull Hypothesis (\(\sf{H_{0}}\)): mpg is not impacted by am_fctr.
The variance by am_fctr appears to be independent. #{r q1, cache=FALSE} # print(t.test(subset(cars_df, am_fctr == "automatic")$mpg, # subset(cars_df, am_fctr == "manual")$mpg, # var.equal=FALSE)$conf) # We reject the null hypothesis i.e. we have evidence to conclude that am_fctr impacts mpg (95% confidence). Manual transmission is better for miles per gallon versus automatic transmission.
## label step_major step_minor bgn end elapsed
## 10 fit.models 7 0 75.172 186.411 111.239
## 11 fit.models 7 1 186.412 244.918 58.506
## 15 fit.data.training 8 1 299.115 345.244 46.129
## 16 predict.data.new 9 0 345.244 382.729 37.485
## 12 fit.models 7 2 244.919 268.677 23.758
## 2 inspect.data 2 0 19.517 42.418 22.901
## 4 encode.data 2 2 45.384 63.659 18.275
## 14 fit.data.training 8 0 281.597 299.115 17.518
## 13 fit.models 7 3 268.677 281.596 12.920
## 8 select.features 5 0 66.701 74.324 7.624
## 1 import.data 1 0 12.518 19.517 6.999
## 3 scrub.data 2 1 42.419 45.384 2.965
## 6 extract.features 3 0 64.137 65.840 1.703
## 7 cluster.data 4 0 65.840 66.701 0.861
## 9 partition.data.training 6 0 74.325 75.172 0.847
## 5 manage.missing.data 2 3 63.659 64.136 0.478
## duration
## 10 111.239
## 11 58.506
## 15 46.129
## 16 37.485
## 12 23.758
## 2 22.901
## 4 18.275
## 14 17.518
## 13 12.919
## 8 7.623
## 1 6.999
## 3 2.965
## 6 1.703
## 7 0.861
## 9 0.847
## 5 0.477
## [1] "Total Elapsed Time: 382.729 secs"
## label step_major step_minor bgn end elapsed
## 2 fit.models_1_rpart 2 0 190.336 244.878 54.542
## 1 fit.models_1_bgn 1 0 190.320 190.336 0.016
## duration
## 2 54.542
## 1 0.016
## [1] "Total Elapsed Time: 244.878 secs"
## R version 3.2.0 (2015-04-16)
## Platform: x86_64-apple-darwin13.4.0 (64-bit)
## Running under: OS X 10.10.3 (Yosemite)
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] tcltk grid parallel stats graphics grDevices utils
## [8] datasets methods base
##
## other attached packages:
## [1] gdata_2.16.1 rpart.plot_1.5.2 rpart_4.1-9 caTools_1.17.1
## [5] dplyr_0.4.1 plyr_1.8.2 sqldf_0.4-10 RSQLite_1.0.0
## [9] DBI_0.3.1 gsubfn_0.6-6 proto_0.3-10 reshape2_1.4.1
## [13] doMC_1.3.3 iterators_1.0.7 foreach_1.4.2 doBy_4.5-13
## [17] survival_2.38-1 caret_6.0-47 ggplot2_1.0.1 lattice_0.20-31
##
## loaded via a namespace (and not attached):
## [1] gtools_3.5.0 splines_3.2.0 colorspace_1.2-6
## [4] htmltools_0.2.6 yaml_2.1.13 mgcv_1.8-6
## [7] chron_2.3-45 e1071_1.6-4 nloptr_1.0.4
## [10] RColorBrewer_1.1-2 stringr_1.0.0 munsell_0.4.2
## [13] gtable_0.1.2 codetools_0.2-11 evaluate_0.7
## [16] labeling_0.3 knitr_1.10.5 SparseM_1.6
## [19] quantreg_5.11 pbkrtest_0.4-2 class_7.3-12
## [22] Rcpp_0.11.6 scales_0.2.4 formatR_1.2
## [25] BradleyTerry2_1.0-6 lme4_1.1-7 digest_0.6.8
## [28] stringi_0.4-1 brglm_0.5-9 tools_3.2.0
## [31] bitops_1.0-6 magrittr_1.5 lazyeval_0.1.10
## [34] car_2.0-25 MASS_7.3-40 Matrix_1.2-1
## [37] assertthat_0.1 minqa_1.2.4 rmarkdown_0.6.1
## [40] compiler_3.2.0 nnet_7.3-9 nlme_3.1-120